Salobo Copper-Gold Mine Salobo III Expansion · assessment, pre-feasibility, and feasibility studies, and due diligence studies in Canada, Brazil, Tanzania, Ghana and Portugal. As

1

Salobo Copper-Gold Mine Carajás, Pará State, Brazil Technical Report – Salobo III Expansion

Neil Burns, P.Geo. Chris Gauld, P.Geo Marcos Dias Alvim, P.Geo., FAusIMM(CP) Maurice Tagami, P.Eng. Effective date: December 31, 2019

CERTIFICATE OF QUALIFIED PERSON

I, Neil Burns, M.Sc., P.Geo., am employed as Vice President, Technical Services, Wheaton

Precious Metals Corp. (Wheaton).

This certificate applies to the technical report titled “Salobo Copper-Gold Mine Carajás, Pará

State, Brazil – Technical Report – Salobo III Expansion” that has an effective date of December

31, 2019 (the “technical report”).

I am a professional geologist with over 24 years of exploration, mining and resource geology

experience in precious and base metals. I graduated from Dalhousie University with a B.Sc in

1995 and from Queen’s University with a M.Sc. in 2003. I have practiced professionally since

graduation in 1995. In that time I have been directly involved in generation of, and review of,

mineral tenure, surface and other property rights, geological, mineralization, exploration and

drilling data, geological models, sampling, sample preparation, assaying, quality assurance-

quality control databases, mineral resource estimation, risk analyses, mine geology,

reconciliation, preliminary economic assessment, pre-feasibility and feasibility studies, and due

diligence studies in Canada, USA, Central and South America, Europe, Eurasia, Africa and

Australia.

As a result of my experience and qualifications, I am a Qualified Person as defined in National

Instrument 43–101 Standards of Disclosure for Mineral Projects (NI 43–101).

I have visited the Salobo Operations, most recently from December 2nd to 4th, 2019.

I am responsible for Sections 1, 2, 3, 7, 8, 9, 10, 11, 12, 14, 22, 23, 24, 25, 26 and 27 of the

technical report.

I am not independent of Wheaton as I am an employee of Wheaton. I have read NI 43–101 and

the sections of the technical report for which I am responsible have been prepared in compliance

with that Instrument.

As of the effective date of the technical report, to the best of my knowledge, information and

belief, the sections of the technical report for which I am responsible contain all scientific and

technical information that is required to be disclosed to make those sections of the technical

report not misleading.

March 30, 2020

“Signed and Stamped”

Neil Burns, M.Sc., P.Geo.

Wheaton Precious Metals Corp.

Suite 3500 - 1021 West Hastings Street

Vancouver, BC Canada V6E 0C3


I, Chris Gauld M.Sc., P.Geo., am employed as Principal Geologist, Resource Management, Vale

Canada Limited.




I am a professional geologist with over 22 years of exploration, underground, open pit and

laterite mining and resource geology experience in base and precious metals. I graduated from

the University of Glasgow, Scotland with a B.Sc (Hons) in 1996 and from Laurentian University

with a M.Sc. in 1999. I have practiced professionally since graduation in 1996. In that time I

have been directly involved in the generation or review of, surface and other property rights,

geological, mineralization, exploration and drilling data, geological models, sampling, sample

preparation, assaying, QA-QC databases, mineral resource estimation, risk analyses, mine

geology, reconciliation, preliminary economic assessment, pre-feasibility and feasibility studies,

and due diligence studies in Canada, Brazil, and Indonesia.



I have visited the Salobo Operations, most recently on October 24-25th, 2019.

I am responsible for Sections 4, 5, 6, 7, 8, 9, 10, 11, 12, 23, 24, 25 and 26 of the technical report.

I am independent of Wheaton as independence is defined in Section 1.5 of NI 43-101 however I

am employed by Vale and therefore not independent of Vale, the owner of Salobo. I have read

NI 43–101 and the sections of the technical report for which I am responsible have been prepared

in compliance with that Instrument.





March 30th, 2020


Chris Gauld, M.Sc., P.Geo.

Vale Base Metals

18 Rink Street

Copper Cliff, ON

P0M 1N0


I, Marcos Dias Alvim, MBA., P.Geo., FAusIMM(CP), am employed as Long Term Planning

Manager, South Atlantic Operations, Vale Base Metals (Vale).




I have worked as a geologist for a total of 12 years since my graduation. I graduated with a B.Sc.

in Geology from the Universidade Estadual Paulista “Julio de Mesquita Filho”, Brazil, in 2007

and a MBA in Project Management from the Fundação Getúlio Vargas, Brazil, in 2013. I have

done review and report work on many mining operations and projects around the world for due

diligence and regulatory requirements. I have participated in scoping and feasibility study

project work on many mining projects, including Brazilian projects. I have held senior positions

at numerous base metal and precious metal open pit and underground operations, and consulting

companies responsible for general and project management, including Manager of Technical

Services, Project Development and Operations of major and junior Canadian mining companies.



I have visited the Salobo Operations, most recently on March 5th, 2020.

I am responsible for Sections 4, 14, 15, 16, 18, 19, 20, 21, 22, 25 and 26 of the technical report.

I am independent of Wheaton as independence is defined in Section 1.5 of NI 43-101, however I

am employed by Vale and therefore not independent of Vale, the owner of Salobo. I have read

NI 43–101 and the sections of the technical report for which I am responsible have been prepared

in compliance with that Instrument.





March 30, 2020


Marcos Dias Alvim, MBA, P.Geo., FAusIMM(CP)

Vale Base Metals South Atlantic Operations

Rua Grajau, 63, Carajás - Parauapebas - Pará – Brasil

CEP: 68.516-000


I, Maurice Tagami, P.Eng., am employed as Technical Ambassador (formerly Vice President,

Mining Operations), Wheaton Precious Metals Corp. (Wheaton).




I am a professional engineer with over 39 years of metallurgical, mineral processing, mining

operations and project development experience in both base and precious metals. I am a member

of the Association of Professional Engineers and Geoscientists of the Province of British

Columbia and hold a Bachelor of Applied Science degree in Metallurgical Engineering from the

University of British Columbia. I have practiced professionally since graduation in 1981. In that

time, I have been directly involved in generation of, and review of, preliminary economic

assessment, pre-feasibility, and feasibility studies, and due diligence studies in Canada, Brazil,

Tanzania, Ghana and Portugal.



I have visited the Salobo Operations, most recently from December 2nd to 4th, 2019.

I am responsible for Sections 13, 17, 25 and 26 of the technical report.

I am not independent of Wheaton as I am an employee of Wheaton. I have read NI 43–101 and

the sections of the technical report for which I am responsible have been prepared in compliance

with that Instrument.





March 30, 2020


Maurice Tagami, P.Eng.

Wheaton Precious Metals Corp.

Suite 3500 - 1021 West Hastings Street

Vancouver, BC Canada V6E 0C3

December 2019 1

Table of Contents 1. Summary .................................................................................................................. 9

2. Introduction ............................................................................................................ 11

2.1. Terms of Reference ...................................................................................... 11

2.2. Qualified Persons .......................................................................................... 11

2.3. Site Visits ...................................................................................................... 12

2.4. Previous Technical Reports .......................................................................... 12

3. Reliance on Other Experts ..................................................................................... 13

4. Property Description and Location ......................................................................... 13

4.1. Location ........................................................................................................ 13

4.2. Regulatory .................................................................................................... 14

4.3. Mineral Tenure .............................................................................................. 14

4.4. Royalties/Mining Taxes ................................................................................. 15

4.5. Social License ............................................................................................... 16

4.6. Comments on Section 4 ................................................................................ 16

5. Accessibility, Climate, Local Resources, Infrastructure and Physiography ............ 17

5.1. Accessibility .................................................................................................. 17

5.2. Climate .......................................................................................................... 17

5.3. Local Resources and Infrastructure .............................................................. 18

5.4. Physiography ................................................................................................ 18


6. History .................................................................................................................... 19

6.1. Production History ......................................................................................... 22

7. Geological Setting and Mineralization .................................................................... 22

7.1. Regional Geology ......................................................................................... 22

7.2. Property Geology .......................................................................................... 24

7.3. Tectonic Setting ............................................................................................ 28

7.4. Metamorphism .............................................................................................. 29

7.5. Alteration ....................................................................................................... 29

7.6. Mineralization ................................................................................................ 30


8. Deposit Types ........................................................................................................ 31

9. Exploration ............................................................................................................. 32

9.1. Geological Mapping ...................................................................................... 32

December 2019 2

9.2. Airborne Gravity Survey ................................................................................ 33


10. Drilling ................................................................................................................. 35

10.1. Drill Methods ................................................................................................. 38

10.2. Core Reception, Handling and Storage ........................................................ 38

10.3. Geological logging ........................................................................................ 39

10.4. Recovery ....................................................................................................... 40

10.5. Collar Surveys............................................................................................... 40

10.6. Downhole Surveys ........................................................................................ 40

10.7. Specific Gravity Determination ...................................................................... 41

10.8. Comments on Section 10 .............................................................................. 41

11. Sample Preparation, Analyses and Security ....................................................... 42

11.1. Sampling Methods ........................................................................................ 42

11.1.1. Drill Core ................................................................................................ 42

11.1.2. Blast Holes ............................................................................................. 42

11.2. Sample Preparation ...................................................................................... 43

11.2.1. Exploration ............................................................................................. 43

11.2.2. Grade Control ......................................................................................... 45

11.3. Sample Analysis ........................................................................................... 46

11.3.1. Exploration ............................................................................................. 46

11.3.2. Grade Control ......................................................................................... 47

11.4. Quality Assurance and Quality Control ......................................................... 48

11.4.1. Blanks .................................................................................................... 50

11.4.2. Duplicates .............................................................................................. 50

11.4.3. Standard Reference Material.................................................................. 51

11.4.4. Quality Control for Blast-holes ................................................................ 52

11.5. Security ......................................................................................................... 52


12. Data Verification .................................................................................................. 53

12.1. Major Mining Studies .................................................................................... 53

12.2. External Audits and Reviews ........................................................................ 53


13. Mineral Processing and Metallurgical Testing ..................................................... 55

13.1. Metallurgical Testwork .................................................................................. 55

December 2019 3

13.1.1. Variability Tests ...................................................................................... 55

13.1.2. High Pressure Grind Roll Trade-off Study (2006) ................................... 59

13.1.3. Mixed Ore Zone Copper Recovery Testwork ......................................... 60

13.2. Recovery Estimates ...................................................................................... 62

13.3. Metallurgical Variability ................................................................................. 63

13.4. Deleterious Elements .................................................................................... 63

13.5. Actual Plant Results versus Budget and Model Projections .......................... 64

13.5.1. Historical Metallurgical Results ............................................................... 64

13.5.2. Historical Plant Availability and Utilization .............................................. 69

13.5.3. Recovery Projections ............................................................................. 70

13.6. Process Plant Optimization ........................................................................... 72


14. Mineral Resource Estimates ................................................................................ 75

14.1. Introduction ................................................................................................... 76

14.2. Geological Interpretation ............................................................................... 77

14.3. Domaining ..................................................................................................... 79

14.4. Statistical Analysis ........................................................................................ 81

14.4.1. Raw Assay Statistics .............................................................................. 81

14.4.2. Compositing ........................................................................................... 82

14.4.3. Domained Composite Statistics .............................................................. 83

14.4.4. Outlier Analysis ...................................................................................... 83

14.5. Continuity Analysis ........................................................................................ 83

14.6. Block Modeling.............................................................................................. 84

14.6.1. Dimensions............................................................................................. 84

14.6.2. Boundary Conditions .............................................................................. 84

14.6.3. Block Estimation ..................................................................................... 85

14.6.4. Classification Coding .............................................................................. 86

14.6.5. Model Validation ..................................................................................... 86

14.7. Resource Model Pit Optimization .................................................................. 88

14.8. Classification of Mineral Resources .............................................................. 89

14.9. Mineral Resource Tabulation ........................................................................ 90


15. Mineral Reserve Estimates .................................................................................. 92

15.1. Mineability and Dilution ................................................................................. 92

December 2019 4

15.2. Pit Optimization ............................................................................................. 92

15.2.1. General Assumptions ............................................................................. 93

15.2.2. Mineral Resource Model ........................................................................ 93

15.2.3. Mining Costs........................................................................................... 93

15.2.4. Processing Costs ................................................................................... 93

15.2.5. Recovery ................................................................................................ 93

15.2.6. Over Head Costs .................................................................................... 94

15.2.7. Refining, Freight, and Royalties ............................................................. 94

15.2.8. Sustaining Capital .................................................................................. 94

15.2.9. Geotechnical Assumptions ..................................................................... 94

15.3. Selective Mining Unit .................................................................................... 94

15.4. Geotechnical Considerations ........................................................................ 95

15.5. Cut-off Grade ................................................................................................ 97

15.6. Mineral Reserve Tabulation .......................................................................... 97

15.7. Reconciliation ............................................................................................. 101

15.7.1. Method ................................................................................................. 101

15.7.2. Results ................................................................................................. 102

15.8. Comments on Section 15 ............................................................................ 104

16. Mining Methods ................................................................................................. 105

16.1. Pit and Phase Designs ................................................................................ 105

16.2. Production Schedule ................................................................................... 106

16.3. Grade Control ............................................................................................. 107

16.4. Mining Equipment ....................................................................................... 108

16.5. Manpower ................................................................................................... 109

16.6. Ore Stockpiles & Waste Disposal ............................................................... 109

16.7. Mine Services ............................................................................................. 110


17. Recovery Methods ............................................................................................ 110

17.1. Process Flowsheet ...................................................................................... 110

17.2. Plant Design ............................................................................................... 112

17.2.1. Salobo I and II ...................................................................................... 112

17.2.2. Salobo III .............................................................................................. 115

17.3. Energy, Water, and Process Materials Requirements ................................ 129

17.4. Plant Equipment and Design Considerations .............................................. 129

December 2019 5

17.5. Process Plant Performance Projections ...................................................... 131


18. Project Infrastructure ......................................................................................... 133

18.1. Roads and Logistics .................................................................................... 133

18.2. Stockpiles and Waste Rock Storage Facilities ............................................ 134

18.3. Site Infrastructure ........................................................................................ 135

18.4. Tailings Storage Facility .............................................................................. 136

18.5. Water Supply .............................................................................................. 137

18.6. Power and Electrical ................................................................................... 137

18.7. Communications ......................................................................................... 137

18.8. Housing ....................................................................................................... 138


19. Market Studies and Contracts ........................................................................... 138


20. Environmental Studies, Permitting and Social or Community Impact ................ 139

20.1. Environmental Management ....................................................................... 141

20.2. Permitting .................................................................................................... 141

20.3. Social and Community Impact .................................................................... 143

20.4. Closure Plan ............................................................................................... 144


21. Capital and Operating Costs ............................................................................. 144

21.1. Capital Cost Estimates ................................................................................ 144

21.2. Operation Cost Estimates ........................................................................... 145

21.2.1. Mine Operating Costs ........................................................................... 146

21.2.2. Process Operating Costs ..................................................................... 146

21.2.3. Other Operating Costs ......................................................................... 146


22. Economic Analysis ............................................................................................ 147

22.1. Key Assumptions ........................................................................................ 147

22.2. Cash Flow Analysis ..................................................................................... 149

22.3. DCF Analysis Results and Sensitivity Analysis ........................................... 150


23. Adjacent Properties ........................................................................................... 151

24. Other Relevant Information ............................................................................... 151

December 2019 6

25. Interpretation and Conclusions .......................................................................... 151

26. Recommendations ............................................................................................ 152

27. References ........................................................................................................ 153

Table of Figures Figure 1 - Location Map ................................................................................................ 14 Figure 2 - Mineral Tenure Layout Plan .......................................................................... 15

Figure 3 - Regional Geology of the Carajás Province ................................................... 24 Figure 4 - Major Lithological Units – Plan View ............................................................. 27 Figure 5 - Major Lithological Units – Vertical Section View ........................................... 27 Figure 6 – Core Photos of Major Lithological Units ....................................................... 28

Figure 7 - Tectonic Setting of the Carajás Region ......................................................... 28 Figure 8 - Copper Mineralization Styles at Salobo ........................................................ 31

Figure 9 - Coincident Magnetic and Gravimetric Anomalies .......................................... 34 Figure 10 - 3D Gravity Inversion with Current >0.5% Cu Block Model Outline .............. 34 Figure 11 - Plan View – Drill Hole Traces ...................................................................... 37

Figure 12 – Infill & Deep Drilling, Section 1200E ........................................................... 38 Figure 13 - Blast-Hole Sampling Pattern ....................................................................... 43

Figure 14 – 2002 – 2003 -Sample Preparation Flowchart ............................................. 44 Figure 15 – 2017 - 2019 -Sample Preparation Flowchart .............................................. 45 Figure 16 - Copper Recoveries in 2003–2004 Variability Testwork ............................... 57

Figure 17 - Gold Recoveries from Variability and LCT Testwork Programs .................. 57 Figure 18 - Derivation of Copper Recovery Projection, 2003–2004 Variability Testwork ...................................................................................................................................... 58 Figure 19 - Derivation of Gold Recovery Projection from Variability Testwork .............. 58

Figure 20 - Testwork with Mixed Ore – Effect of Dispersant and Collector Dosage ...... 61 Figure 21 - Expected Metallurgy of Various Sulphide-Mixed Ore Blends ...................... 61

Figure 22 - Test Results with Fresh and Mixed Ores with Modified Reagent Scheme .. 62 Figure 23 - Actual versus Projected Monthly Plant Copper Recovery ........................... 66 Figure 24 - Actual versus Projected Monthly Plant Gold Recovery ............................... 66 Figure 25 - Historical Plant Performance – Cu Recovery .............................................. 67

Figure 26 - Historical Plant Performance – Au Recovery .............................................. 67 Figure 27 - Historical Plant Performance – Concentrate Cu Grade ............................... 68 Figure 28 - Historical Plant Performance – Concentrate Au Grade ............................... 68 Figure 29 - Salobo Plant Historical Availability .............................................................. 69 Figure 30 - Salobo Plant Historical Operational Utilization ............................................ 69

Figure 31 - Salobo Line I and II Mineral Processing Facility .......................................... 72 Figure 32 - Salobo Flow Sheet ...................................................................................... 73

Figure 33 - Salobo Deposit Sectors............................................................................... 78 Figure 34 - Cu Grade Shell Model – Section 1100SE (looking west) ............................ 79 Figure 35 - Domain Definition for Copper ...................................................................... 80 Figure 36 - Domain Definition for Gold .......................................................................... 81 Figure 37 - Domain 1203 Swath Plot - Copper .............................................................. 87 Figure 38 - Domain 1203 Swath Plot - Gold .................................................................. 87

December 2019 7

Figure 39 – Mineral Resource Pit – Isometric View NNE .............................................. 89

Figure 40 – Geotechnical Design Sectors ..................................................................... 96

Figure 41 - 2019 MRMR Longitudinal Section (looking southwest) ............................. 100 Figure 42 - Production Reconciliation Schematic ........................................................ 101 Figure 43 – Salobo Phases ......................................................................................... 105 Figure 44 - Life of Mine Plan (Mineral Reserves Only) ................................................ 107 Figure 45 - Simplified Process Flowsheet ................................................................... 113

Figure 46 - Mine site layout with Salobo III .................................................................. 116 Figure 47 - Salobo III Simplified Process Flowsheet ................................................... 117 Figure 48 - Salobo III Flotation Circuit ......................................................................... 122 Figure 49 - Process Plant Single Line Equipment ....................................................... 131 Figure 50 - Concentrate Load Out, Parauapebas ....................................................... 134

Figure 51 - Location Plan, Waste Rock Storage Facility and Low-Grade Stockpiles .. 135

Figure 52 - Salobo Tailings Facility with 2019 Lift Completed ..................................... 136

Figure 53 - Tailings Storage Capacity and Area versus Impoundment Height ............ 137 Figure 54 - Mine Location in Relation to Forest Reserves ........................................... 140

Figure 55 - NPV Sensitivity Analysis ........................................................................... 151

Table of Tables Table 1 - December 31, 2019 Mineral Reserves & Mineral Resources ......................... 10 Table 2 - Salobo Production .......................................................................................... 22

Table 3 - Exploration Summary 1978–2003 .................................................................. 32 Table 4 - Drill Hole Summary Table .............................................................................. 36

Table 5 - Blast-Hole Sample Analysis ........................................................................... 47

Table 6 - Adjustment for Copper Assays for pre-2002 Drilling Programs ...................... 48

Table 7 - Adjustment for Gold Assays for pre-2002 Drilling Programs .......................... 49 Table 8 - QA/QC Sample Types and Insertion Frequency 2017 - 2019 ........................ 50

Table 9 - Summary Duplicate Analysis .......................................................................... 51 Table 10 – 2017 – 2019 Standard Certified Values ....................................................... 52 Table 11 - Processing Recovery Assumptions (2020-2023) ......................................... 70 Table 12 - Processing Recovery Assumptions (2024-LoMP) ........................................ 70

Table 13 - Summary of Salobo Infill Diamond Drilling Campaign .................................. 75 Table 14 - December 31, 2019 Mineral Resource Estimates ........................................ 76 Table 15 – Block Model Domains Codes ...................................................................... 80 Table 16 - Block Model Zone Codes ............................................................................. 80 Table 17 - Copper Assay Statistics by Lithology ........................................................... 82

Table 18 - Gold Assay Statistics by Lithology ............................................................... 82

Table 19 - Grade Capping Levels ................................................................................. 83

Table 20 – Outlier Restriction ........................................................................................ 83 Table 21 – Salobo Variography ..................................................................................... 84 Table 22 - Block Model Origin ....................................................................................... 84 Table 23 - Boundary Conditions .................................................................................... 85 Table 24 - Search Ellipse Dimensions – Copper & Gold ............................................... 85 Table 25 - Global Mean Analysis .................................................................................. 86 Table 26 - Mineral Resource Open Pit Optimization Assumptions ................................ 88

December 2019 8

Table 27 - Mineral Resources from 2018 to 2019 ......................................................... 90

Table 28 - Changes to Mineral Resources from 2018 to 2019 ...................................... 91

Table 29 – Geotechnical Design Sectors for Salobo Mine ............................................ 95 Table 30 – Cut-off Calculation Parameters ................................................................... 97 Table 31 - December 31, 2019 Mineral Reserves ......................................................... 98 Table 32 - Summary of Mineral Reserves from 2018 to 2019 ....................................... 98 Table 33 - Changes to Mineral Reserves from 2018 to 2019 ........................................ 99

Table 34 - 2019 Mineral Reserves Estimate by Phase................................................ 100 Table 35 – 2018 / 2019 Calculated Reconciliation Factors ......................................... 103 Table 36 - Mill versus Production ................................................................................ 104 Table 37 - December 31, 2019 Mineral Reserve Estimate by Phase .......................... 106 Table 38 - 5 Year Plan ................................................................................................ 106

Table 39 - Actual Production Mining Fleet ................................................................... 109

Table 40 - Waste Dump & Stockpile Design Parameters ............................................ 109

Table 41 - Major Process Equipment .......................................................................... 130 Table 42 - Processing Recovery Assumptions (2020-2023) ....................................... 131

Table 43 - Processing Recovery Assumptions (2024-LoMP) ...................................... 132 Table 44 - Processing Plant Performance Forecasts .................................................. 132

Table 45 - Actual and Forecasted Processing Recoveries .......................................... 132 Table 46 - Metal Sale Price and Exchange Rate Assumptions ................................... 139 Table 47 - Historical Capital Expenditures Summary (US$ M) .................................... 144

Table 48 - Capital Expenditures Summary (US$ M).................................................... 145 Table 49 - Historic Operating Costs (US$ M) .............................................................. 145

Table 50 - Forecast Operating Cost ............................................................................ 146 Table 51 - Mine Operating Costs ................................................................................ 146 Table 52 - Process Operating Costs ........................................................................... 146

Table 53 - Other Costs ................................................................................................ 147

Table 54 - Metal Sale Price and Exchange Rate Assumptions ................................... 148 Table 55 - Cash Flow (Reserves only) ........................................................................ 149 Table 56 - Annual Free Cash Flow (Reserves Only) ................................................... 150

Table 57 - NPV – Mineral Reserves Only .................................................................... 150

December 2019 9

1. Summary The Salobo Operations comprise a large open pit mine and concentrator facilities located in the Carajás Mining District in eastern Pará State, Brazil. The mine is 100% owned by Vale S.A. This technical report (the Report) summarizes the supporting information for the December 31, 2019 Mineral Reserve and Mineral Resource estimates for the Salobo Operations and summarizes the material information relating to the Salobo Operations. The Report has been prepared in accordance with the requirements of National Instrument 43-101 Standards of Disclosure for Mineral Projects (NI 43-101) and Form 43–101F1. All estimates are supported by applicable scientific and technical information. Unless otherwise indicated, all financial values are reported in US currency and metric as the units of measure. The Salobo deposit is hosted in the Carajás Mining District within Carajás Province, a sigmoidal-shaped, west–northwest to east–southeast-trending late Archean basin. The basin contains a basement assemblage that is dominated by granite–tonalitic ortho-gneisses of the Pium Complex, and amphibolite, gneisses and migmatites of the Xingu Complex. The basement rocks are overlain by volcanic and sedimentary rocks of the Itacaiúnas Supergroup, which includes the Igarapé Salobo Group, the Igarapé Pojuca Group, Grão Pará Group and the Igarapé Bahia Group. The Itacaiúnas Supergroup hosts all the Carajás iron deposits as well as Salobo. Salobo is considered to be an example of an iron oxide–copper–gold (IOCG) deposit. Global examples include Olympic Dam in Australia, Candelaria–Punta del Cobre in Chile, and Sossego in Brazil. The major host units are biotite and magnetite schists. The Salobo hydrothermal system has a core of massive magnetite that is surrounded by less intensely altered rocks. Away from the massive magnetite, the magnetite content gradually diminishes, giving way to biotite–garnet schist and / or garnet–grunerite schist. Sulphide mineralization typically consists of assemblages of magnetite–chalcopyrite–bornite and magnetite–bornite–chalcocite. Copper mineralization was discovered by a Vale predecessor company in 1974 and detailed exploration commenced in 1977. Initial exploration efforts included stream sediment sampling, reconnaissance exploration, and ground induced polarization (IP) and magnetometer geophysical surveys. Follow-up work in 1978 identified the presence of copper sulphides in an outcrop of magnetite schists at Salobo. Core drilling commenced in 1978 and was conducted through to 2003 in five different drilling campaigns. An infill drilling program was initiated in 2017, the first core drilling since 2003.

December 2019 10

A scoping study was completed in 1981, and pilot studies ran from 1985 to 1987, culminating in the grant of a mining concession. A prefeasibility study was concluded in 1988, an initial feasibility study was conducted in 1998, updates to the feasibility study were undertaken in 2001 and 2002, and a final study was completed in 2004. The Salobo Mine commenced pre-stripping in 2009. Project ramp-up for Phase I (12 Mtpa) of the Salobo Operations was completed three years later and the first concentrate was shipped in September 2012. Phase II, doubling the nameplate capacity, was completed in 2014. In 2019 Vale commenced construction of a phase III expansion (Salobo III) which will increase production from 24 Mtpa to 36 Mtpa. Initial production from the expansion is expected in January 2022. Table 1 details the Salobo Operations Mineral Reserves and Mineral Resources as of December 31, 2019.

Table 1 - December 31, 2019 Mineral Reserves & Mineral Resources

Item Classification M Tonnes Cu % Au g/t

Mineral Reserves

Proven 152.7 0.69 0.39

Probable 832.4 0.62 0.32

Stockpiles (Proven) 163.4 0.45 0.22

P&P 1,148.4 0.60 0.32

Mineral Resources within pit

Measured

Indicated

M&I

Inferred 21.9 0.50 0.20

Mineral Resources adjacent to pit

Measured 1.2 0.67 0.42

Indicated 192.3 0.61 0.31

M&I 193.5 0.61 0.31

Inferred 154.1 0.60 0.30

Total Mineral Resources

Measured 1.2 0.67 0.42 Indicated 192.3 0.61 0.31

M&I 193.5 0.61 0.31

Inferred 176.1 0.59 0.29

Notes: 1. Mineral Resource estimates were prepared by Mr. Joao Dirk V. Reuwsaat and Mineral Reserve estimates by Mr. Wellington

F. de Paula, both Vale employees. The Qualified Person for the Mineral Resource and Mineral Reserve estimates is Mr. Marcos Dias Alvim, P.Geo., FAusIMM(CP), Long Term Planning Manager, South Atlantic Operations, Vale Base Metals.

2. Mineral Resources are exclusive of Mineral Reserves 3. Mineral Resources that are not Mineral Reserves do not have demonstrated economic viability 4. Mineral Reserves and Mineral Resources are reported above a copper equivalent cut-off of 0.253%, assuming $1,290 per

ounce gold and $3.18 per pound copper

In the opinion of the QPs, the exploration and diamond drilling data were completed according to 2003 CIM Best Practice Guidelines and the Mineral Resources and Mineral Reserves have been estimated according to 2014 CIM Definition Standards. The Salobo Operations is a fully developed mine site with all the required permits and infrastructure. The mine has a large Mineral Reserve base and strong economic margins which result in forecasted mining until the year 2044 and then the processing of stockpiled material until 2052.

December 2019 11

2. Introduction Wheaton Precious Metals Corp (Wheaton) in collaboration with Vale SA, has elected to prepare an updated technical report (the Report), in accordance with the requirements of National Instrument 43-101 Standards of Disclosure for Mineral Projects (NI 43-101) and Form 43–101F1, on the Salobo open pit mining operations (Salobo Operations), located in northern Brazil, in the southeastern portion of Pará State. This report provides and overview of the Salobo Operations and an operational update, including Mineral Reserves and Mineral Resources and detail the Salobo III expansion which is in construction at the time of this report. The corporate entity that conducts the mining operations is Salobo Metais SA (SMSA), an indirectly wholly-owned subsidiary of Vale SA. For the purposes of this Report, unless otherwise noted, Vale SA and Salobo Metais SA will be referred to interchangeably as Vale. Wheaton’s interest in the Salobo Operations is restricted to a metal streaming agreement that applies to 75% of the gold produced as a byproduct at the Salobo Operations for the life of the mine (the streaming agreement). The Salobo Operations consist of an operating copper–gold open pit mine, currently producing at a rate of 24 Mtpa through a conventional crush–grind–float processing plant, producing copper concentrates. The Salobo III expansion, when completed, will add 12 Mtpa to the processing capacity for a total of 36 Mtpa.

2.1. Terms of Reference The Report was prepared to support scientific and technical disclosure on the Salobo Operations in Wheaton’s Annual Information Form for the year ending 31 December, 2019. Mineral Resources and Mineral Reserves are reported with reference to the Canadian Institute of Mining, Metallurgy and Petroleum (CIM) Definition Standards for Mineral Resources and Mineral Reserves (May 2014; the 2014 CIM Definition Standards) and the CIM Estimation of Mineral Resources and Mineral Reserves Best Practice Guidelines (November 2019; 2019 CIM Best Practice Guidelines). All measurement units used in this Report are metric units and currency is expressed in US dollars (US$), unless stated otherwise. The Brazilian currency is the Real (BRL). The Report uses Canadian English.

2.2. Qualified Persons The following Qualified Persons (QPs), as defined in NI 43-101 authored this Report:

December 2019 12

Wheaton:

• Neil Burns, P.Geo., Vice President, Technical Services, Wheaton

• Maurice Tagami, P.Eng., Technical Ambassador, Wheaton Vale:

• Chris Gauld, P.Geo., Principal Geologist, Resource Management, Vale Base Metals

• Marcos Dias Alvim, P.Geo., FAusIMM(CP), Long Term Planning Manager, South Atlantic Operations, Vale Base Metals

2.3. Site Visits Mr. Neil Burns and Mr. Maurice Tagami make annual visits to the Salobo Operations to review developments. Mr. Burns’ focus is on Mineral Resources and Mineral Reserves and Mr. Tagami’s focus is on mineral processing. However, both are involved in a full site review of operational performance including discussions with site personnel on aspects of exploration, metallurgy, infrastructure, workforce, mining, equipment, budgeting, environmental, tailings, social considerations and permitting. Both have been to site numerous times since 2013 when the first Wheaton streaming deal was completed, most recently December, 2nd to 4th, 2019. Since 2016, Mr. Chris Gauld has made at a minimum, annual visits to the Salobo Operations to review all matters pertaining to Mineral Resources and Mineral Reserves, exploration and mining. Mr. Gauld’s most recent site visit was October 24th to 25th, 2019. Mr. Marcos Dias Alvim has made at minimum, monthly visits to the Salobo Operations to review all aspects related to exploration, mine geology, mining and Mineral Resources and Mineral Reserves estimation since September 2019. Mr. Alvim’s last visited the mine on March 5, 2020.

2.4. Previous Technical Reports Wheaton has previously filed the following reports on the Salobo Operations: Osmond, J.C., Foo, B., Turner, J., and Jacobs, C., 2013: Technical Report on the Mineral Reserves and Mineral Resources of the Salobo Copper-Gold Mine Carajás, Pará State, Brazil: technical report prepared by Micon Inc. for Wheaton, effective date 31 December 2012. Vos, G., Verly, G., Simon, A., Lacombe, P., Hickson, D., Khera, V., and Searston, S., 2016: Salobo Operations Para State, Brazil NI 43-101 Report: technical report prepared by AmecFW for Wheaton, effective date 31 December 2015.

December 2019 13

Burns, N., Davis, C., Diedrich, C., Tagami, M., 2018: Salobo Copper-Gold Mine Carajas, Para State, Brazil Technical Report, effective date 31 December 2017.

3. Reliance on Other Experts The QPs have not relied on any other experts.

4. Property Description and Location

4.1. Location The Salobo operation is located along the southern margin of the Amazon Basin, northern central Brazil, in the southeastern part of the State of Pará (Figure 1). It is also located in the Parauapebas micro-region in the municipality of Marabá and is part of the Carajás Mineral Province. Geographic coordinates for the operation are 5°47’25” S latitude and 56°32’5” W longitude. In addition to major iron mines, the Carajás Mineral Province also hosts manganese deposits, gold mines, copper sulphide and nickel laterite resources. The Carajás area has excellent infrastructure including the all-weather commercial airport at Carajás, which operates with 737 Boeing type aircraft, electrical power derived from the Tucuruí Dam, abundant water, good roads, and social institutions. The Salobo Operations is a copper-gold deposit located approximately 80 km northwest of Carajás, Pará State in northern Brazil. The area is well-served by railroads and highways that connect the villages and cities. Air service is available at the Carajás airport, which is approximately 70 km from Salobo and is capable of receiving commercial aircrafts and it is served by two daily flights to Belém (Pará state major’s city) and to the main Brazilian cities. Marabá is approximately 240 km from Salobo Operations by highway.

December 2019 14

Figure 1 - Location Map

4.2. Regulatory During the course of Vale Base Metals operations at Salobo, Vale is subject to routine claims and litigation incidental to Vale’s business as well as various environmental proceedings. For greater certainty, none of these ongoing legal issues are considered to pose any “moderate”, “major” or “catastrophic” legal risk to the company’s ability to exploit the Mineral Reserves / Mineral Resources reported in this Report. The operation has secured all material licenses and permits; the company has secured all requisite mineral rights and surface rights; there are no material issues of non-compliance that may impact the company’s ability to exploit the resource; and there is no litigation that may impact the company’s ability to exploit the resource. Vale holds clear mineral title to the deposit areas and has all the necessary permits for operation of the mine.

4.3. Mineral Tenure The Salobo Operations tenement title is 100% owned by Vale S.A. The Salobo Operations are located on one claim. The area named Salobo (copper ore, DNPM 807.426/74) refers to Exploration Permit no. 1121 that is dated July 14, 1987, and defined as a polygon of 9,180.61 ha (Figure 2). There was no change to the land tenement status in 2019.

December 2019 15

Brazilian legislation separates the ownership of the surface rights from mineral ownership. A mining company can operate a mine even if does not own the surface, provided it owns the minerals. In this case it is necessary to pay a royalty to the surface owner. The royalty is calculated as 50% of the CFEM (Compensation for Financial Exploitation of Mineral Resources), which is paid to the government. The mining concessions are updated every year on presentation by Vale of the annual report of mining production to the DNPM.

Figure 2 - Mineral Tenure Layout Plan

Note: Figure courtesy Vale, 2019.

4.4. Royalties/Mining Taxes The Compensação Financeira pela Exploração de Recursos Minerais (CFEM) was enacted by legislation in 1989 and is based on a percentage of the holder’s net profit. The value of CFEM varies from 0.2 to 3.0% of the net sales of mineral products:

• 3.0%: aluminum ore, manganese, rock salt and potassium

• 2.0%: iron ore, fertilizer, coal and other substances

• 0.2%: precious stones, colored gemstones, carbonates and noble metals

• 1.0%: gold. The majority of minerals incur the 2.0% royalty.

December 2019 16

Of the amount collected, 65% is paid to the municipalities where production takes place, 23% is paid to the host state, and 12% to the Federal government. Since 2013, Wheaton has entered into the following three different life of mine gold stream agreements on Salobo with Vale, each for 25%, for a total of 75%. In each of the agreements Wheaton agreed to ongoing payments of the lesser of $400 (subject to a 1% annual inflation adjustment now commencing in 2019 on the entire 75% stream) and the prevailing market price for each ounce of gold delivered under the agreement.

• February 2013 - Vale entered into an agreement with Wheaton to sell 25% of the gold produced at the Salobo Operations for the life of the mine. In exchange, Vale received an initial cash payment of $1.33 billion and 10 million share purchase warrants exercisable at a strike price of $65 per common share.

• March 2015 - Wheaton acquired an additional 25% of the gold production, increasing the gold stream to 50%. Under the amended 2015 streaming agreement, Wheaton paid Vale a cash consideration of $900 million for the new gold stream.

• August 2016 – Wheaton agreed to acquire an additional 25% of the life of mine gold production from the Salobo Operations. This acquisition was in addition to the 50% of the Salobo gold production that Wheaton was entitled to. Wheaton paid upfront cash consideration of $800 million for the increased gold stream and the 10 million Wheaton common share purchase warrants previously issued to a subsidiary of Vale were amended to reduce the strike price from US$65.00 to US$43.75 per common share.

4.5. Social License Areas reserved for indigenous populations are designated as “restricted access” or “prohibited” access for mining. The Brazilian Constitution requires that any mining activities in indigenous areas requires prior approval of the Brazilian National Congress. Indigenous communities have the right to receive royalties from any mining in their areas. In addition to the indigenous communities, there are other communities (Quilombolas) that have Constitutional rights to own and occupy specific lands. Mining is permitted in these areas; however, the communities are entitled to compensation, and if the community needs to be relocated for mining purposes, the community must be relocated to land that has similar characteristics to the area that was previously occupied or be fairly compensated.

4.6. Comments on Section 4 In the opinion of the QPs, the information discussed in this section supports the declaration of Mineral Resources and Mineral Reserves, based on the following:

December 2019 17

• Vale’s mining tenure held is valid and is sufficient to support estimation of Mineral Resources and Mineral Reserves.

• Vale holds sufficient surface rights in the Project area to support the mining operations envisaged in the life-of-mine plans, including access and power line easements.

• Vale currently holds the appropriate permits under local, Provincial and Federal laws to allow mining operations (refer to Section 20). Some permits will require renewal over the course of the planned life-of-mine.

• The appropriate environmental permits have been granted (refer to Section 20)

• At the effective date of this Report, environmental liabilities are typical of an operating open pit mine (refer to Section 20)

• Vale is not aware of any significant environmental, social or permitting issues that would prevent continued exploitation other than those discussed in the Report.

• There is no active artisanal mining on or near the property.

• To the extent known, there are no other significant factors and risks known to Vale that may affect access, title, or the right or ability to perform work at Salobo Operations.

5. Accessibility, Climate, Local Resources, Infrastructure and Physiography

5.1. Accessibility Mining is the primary industry in the area. The Salobo Operations are connected via an all-weather road network to the cities of Parauapebas (80 km), Marabá (240 km), and the commercial airport at Carajás. The Carajás airport can accommodate large aircraft and is served by daily flights to Belém (Pará State major’s city) and other major Brazilian cities. Rail lines carry Salobo and Sossego copper concentrate and iron ore from Carajás to the port city of São Luis. Concentrate from the mine is hauled by trucks to a rail-loading site, north of Parauapebas. Concentrate is then transported, approximately 870 km, by train to the port of São Luís.

5.2. Climate The operations are located in the Carajás mountain range in the eastern Amazon humid tropical rainforest. Temperatures range from 20.8°C to 37.8°C with an average relative humidity of 80.5%. Mean annual rainfall is 1,920 mm and evaporation is 1,500 mm. Winds are predominantly from the north and west. Mining operations are conducted year-round.

December 2019 18

5.3. Local Resources and Infrastructure Mining is the primary industry of the area. As well as Salobo, Vale also operates the Sossego copper mine, located 136 km by road to the south of Salobo and the very large iron ore mine at Carajás located 50 kilometres south east of the mine. Local housing is available for employees within the communities surrounding the mine. There are adequate schools, medical services and businesses to support the work force. The mine site has medical facilities to handle emergencies. In addition, medical facilities are available in Carajás to support the mine’s needs. Vale has invested significantly in infrastructure at Carajás, building a 130 km paved road to Parauapebas and a 20 km sewage system, together with a school, hospital, and day care center. Project infrastructure and the infrastructure layout are discussed in detail in Section 18 of the Report.

5.4. Physiography Salobo is in the northwest of the Carajás Reserve within the 190,000 ha Flona de Tapirapé–Aquiri forest. The area is heavily forested and dominated by relatively dense trees with substantial underbrush. In the mine area, the topography is fairly steep, varying between 190 to 520 m in elevation. The ridge where the Salobo deposit is located has a nominal slope of 2.5H:1.0V. The site is lower than the Carajás Ridge, which is 850 m above sea level. The two drainages on either side of the Salobo Ridge are the Cinzento and Salobo Igarapés (small rivers) which flow into the Itacaiúnas River. The Itacaiúnas River flows into the Tocantins River close to Marabá City. The long-term average unit runoff for the Project site is 13.5 L/s/km2.

5.5. Comments on Section 5 In the opinion of the QPs:

• All necessary infrastructure has been built on site, is operational, and is sufficient for the projected LoMP (see also Section 18).

• There is sufficient suitable land available within the mineral tenure held by Vale for tailings disposal, mine waste disposal, and installations such as the process plant and related mine infrastructure (see also Section 18).

December 2019 19

6. History

• 1974 - CVRD (Companhia Vale do Rio Doce, a predecessor company to Vale) discovered copper mineralization in the Igarapé Salobo region and commenced detailed exploration in 1977. Work completed included stream sediment sampling, reconnaissance exploration, and ground induced polarization (IP) and magnetometer geophysical surveys. As a result, various targets were identified.

• 1978 - The 1974 Salobo exploration targets were revisited and the presence of copper sulphides in an outcrop of magnetite schists at the Salobo 3 Alfa target was noted. Drilling of this target followed in conjunction with the development of two exploration adits. The Salobo 3 Alfa target is now referred to as Salobo.

• 1978 to 1983 - Drilling was initially conducted on a 400 m by 200 m drill grid, subsequently reduced to 200 m by 200 m, and then to 200 m by 100 m. A total of 65 core drill holes (29,322 m) were drilled between March 1978 and May 1983.

• 1981 - A preliminary assessment of potential Project economics was performed, based on an initial resource estimate. The findings were encouraging, and the Carajás Copper Project team submitted an Exploitation Economical Plan for the Salobo deposit to the DNPM in June 1981.

• 1985 – 1987 - A pilot-scale study was carried out from 1985 to 1987 to further define the mineralization style and geometry. This included additional drilling and an additional 1 km of exploration adits. A second drill campaign ran from January 1986 to June 1987. The grid spacing in the core of the deposit was reduced to 100 m by 100 m. Additional drilling was undertaken in the southeast of the deposit from the G-3 adit. This phase included 9,033 m of diamond drilling from 60 drill holes.

• 1987 - The MME granted CVRD mining rights through Ordinance No. 1121.

• 1988 - A prefeasibility study was completed by Bechtel.

• 1993 - Salobo Metais S.A. was incorporated on 29 June 1993 as a joint-venture vehicle between CVRD and Morro Velho Mining (a subsidiary of Anglo American Brasil Ltda. AABL). A third drill campaign was initiated. The primary objective was to investigate the best probable location in the deposit in which to commence mining and to optimize the first five years of production, as well as to investigate mineralized continuity at depth.

• 1993 to 1994 - A total of 64 drill holes (14,585 m) were completed.

• 1997 - A fourth drilling campaign was conducted, resulting in 25,491 m in 88 holes. Mineral Resources Development Inc. (MRDI) audited the drilling information that year.

December 2019 20

• 1998 - A feasibility study was undertaken by Minorco.

• 2001 – The feasibility study was revised and updated by Kvaerner.

• 2002 - AMEC audited the drilling, sampling, assaying and databases that supported the Kvaerner study.

o Changes were made to the Exploitation Economic Plan allowing Salobo Metais to extract silver and gold were approved by DNPM. The original authorization had been for copper only.

o In June 2002, the Brazilian Council for Economic Defense (Conselho

Administrativo de Defesa Econômica) approved the acquisition by CVRD of the 50% of Salobo Metais that was held by AABL. CVRD thus became the largest shareholder in Salobo Metais.

o A fifth drilling campaign drilled 133 drill holes (66,243 m)

• 2003 - A further 2,047 m of drilling was completed, and some areas were drilled at a closer spacing of 50 m x 50 m, including the area around the G3 adit.

• 2006 – Final Pre-Feasibility Study and Installation Licence Granting.

• 2007 – Final Feasibility Study and construction start-up of Salobo I (12Mtpa).

• 2009 - Commenced pre-stripping.

• 2010 – Construction start-up of Salobo II (24Mtpa).

• 2012 - Project ramp-up for Phase I of the Salobo Operations was completed and the first concentrate was shipped in September 2012.

• 2013 – The first Wheaton streaming deal was completed for 25% of the life of mine gold production

o December 2013, the plant processed 898,000 t of ore, which represented 90% of the Phase I nameplate capacity (1 Mt run-of-mine (ROM) per month).

• 2014 - Phase II, intended to double the nameplate capacity and was completed.

• 2015 – The second Wheaton stream deal completed for an additional 25% of the life of mine gold production, increasing the total stream to 50%.

December 2019 21

• 2016 – The third Wheaton stream deal completed for an additional 25% of the life of mine gold production, increasing the total stream to 75%.

• 2017 - During 2017, the following important changes occurred at the Salobo Operations:

o The production data reconciliation process was revised and updated.

o A medium range definition diamond drilling campaign was started.

o A deep exploration drill hole was started to investigate the orebody below

the final pit design.

o The mine and plant quality control (sampling, etc.) process was externally audited.

o A short-term deleterious estimation process for carbon, uranium, fluorine,

sulphur and chlorine was started.

o The phases/pushback design were modified together with the mining plan revision, changing from seven to eight phases.

• 2018 - During 2018, the following important changes occurred at the Salobo Operations:

o The infill drilling program for long-range planning ramped up and approximately 25,000 m was drilled since 2017.

o Three deep exploration drill holes were drilled to investigate the orebody

below the final pit design.

o The GDMS database system was implemented at Salobo to improve the drill core logging process and database security.

• 2019 - During the past 12 months, the following important changes occurred at the Salobo Operations:

o The infill diamond drilling program for long-range planning is running according to plan and approximately 35,000 m has been drilled since 2017. All the sampling and analyses backlog from 2017 and 2018 was completed.

o The fourth deep exploration drill hole below the planned pit was completed with a total length of 1,400 m.

December 2019 22

o The GDMS database system that was implemented for long-term geology last year, was expanded to short-term geology in 2019, improving time and avoiding errors in the data transfer process since the all the analyses results are directed received from the lab equipment.

o Approved the construction of a new core shed for Salobo, in Parauapebas.

o Construction began on the Salobo III Project which consists of a new

beneficiation line with processing capacity of 12 Mtpa and supporting infrastructure. Start-up is scheduled for January 2022.

6.1. Production History Production since mine start-up in 2012 is summarized in the following table:

Table 2 - Salobo Production

Year Feed Concentrate

Tonnage (kt) Cu (%) Au (g/t) Tonnage (t) Cu (%) Au (g/t)

2012 1,823 1.13 0.74 32,231 40.8 20.42

2013 7,366 1.09 0.76 165,471 39.4 21.92

2014 12,474 0.97 0.62 255,511 38.5 19.51

2015 20,290 0.88 0.57 402,592 38.6 19.41

2016 21,401 0.94 0.67 445,238 39.5 22.18

2017 23,650 0.95 0.67 498,172 38.8 21.63

2018 23,657 0.95 0.66 509,811 37.8 22.05

2019 22,486 0.97 0.68 509,778 37.2 22.47

7. Geological Setting and Mineralization

7.1. Regional Geology The Carajás Mining District, located in the southeast of Pará State, lies between the Xingu and Tocantins/Araguaia Rivers, and covers an area of about 300 km x 100 km. It is hosted in the Carajás Province, forming a sigmoidal-shaped, west–northwest to east–southeast-trending late Archean basin (Figure 3). The Archean basin contains a basement assemblage that is dominated by granite–tonalitic ortho-gneisses of the Pium Complex, and amphibolite, gneisses and migmatites of the Xingu Complex. The basement assemblage defines a broad, steeply dipping, east–west trending ductile shear zone (Itacaiúnas shear zone) that experienced multiple episodes of reactivation during the Archean and Paleoproterozoic.

December 2019 23

The metamorphic rocks are cut by Archean-age intrusions, including the calc-alkaline Plaquê Suite (2.73 Ga), and the alkaline Salobo and Estrela granites (2.57 Ga and 2.76 Ga respectively). The basement rocks are overlain by volcanic and sedimentary rocks of the Itacaiúnas Supergroup (2.56 Ga to 2.77 Ga). The Itacaiúnas Supergroup is informally sub-divided as follows (oldest to youngest):

• The Igarapé Salobo Group: iron-rich sediments, quartzites and gneisses, metamorphosed to amphibolite facies; associated with copper–gold and copper–gold–silver mineralization, e.g. Salobo.

• Igarapé Pojuca Group: basic to intermediate volcanic rocks (frequently with cordierite–anthophyllite alteration), amphibolites, gneisses and chemical sediments (cherts), banded iron formation (BIF), and chert; associated with copper–zinc deposits, e.g. Pojuca.

• Grão Pará Group: basal Parauapebas Formation, comprising bimodal volcanic rocks with various degrees of hydrothermal alteration, metamorphism and deformation; upper Carajás Formation, associated with various iron deposits, including the Carajás deposits.

• Igarapé Bahia Group: mafic volcanics (lavas, tuffs and breccias), meta-sediments and BIF, associated with copper, copper–iron, copper–gold–silver deposits, e.g. Igarapé Bahia, Alemão/Bahia and Serra Pelada.

The Itacaiúnas Supergroup hosts all the Carajás iron ore–copper–gold (IOCG) deposits, including Salobo and Sossego, and is thought to have been deposited in a marine rift environment. The metamorphism and deformation has been attributed to the development of a sinistral strike-slip ductile shear zone (the 2.7 Ga Itacaiúnas Shear Zone) and to sinistral, ductile–brittle to brittle transcurrent fault systems. The Itacaiúnas Supergroup is overlain by an extensive succession of Archean marine to fluvial sandstones and siltstones known as the Rio Fresco Group or the Águas Claras Formation (2.68 Ga to 2.78 Ga). The non-deformed, Proterozoic Gorotire Formation, consisting of coarse arkoses and conglomerates with quartz, BIF, and basic rock clasts, overlies the older lithological units (Matos da Costa, 2012). A Proterozoic suite (1.88 Ga) of anorogenic, alkaline granites, the Serra dos Carajás, the Cigano and the Pojuca granites, as well as several generations of younger mafic dykes, cross-cut the entire sequence.

December 2019 24

Figure 3 - Regional Geology of the Carajás Province

Note: Figure modified by AmecFW after SMSA/CVRD, 2003a. The Salobo and Sossego operations (indicated in red) are held by

Vale, as are the 118 and Cristalino prospects/deposits.

7.2. Property Geology Mineralization at the Salobo deposit is hosted by upper greenschist to lower amphibolite metamorphosed rocks of the Igarapé Salobo Group. The group thickness varies from 300–600 m in the Project area, and may be weathered to depths of 30–100 m. The rocks strike approximately N70°W and have a subvertical dip. The major host units are biotite (BDX) and magnetite schists (XMT). Granitic intrusions (GR) occur adjacent to the north and southern sides of the BDX and XMT, and a series of much younger diorite dykes (DB) cross-cut the mineralization forming barren zones. Lithological descriptions of the major units are as follows and as shown in the plan and section views in Figure 4 and Figure 5 respectively and core photos in Figure 6: Magnetite Schist (XMT)

XMT is represented by massive, foliated and banded rocks, with predominant magnetite, fayalite, grunerite, almandine and secondary biotite. Granoblastic textures with polygonal contacts in magnetite and fayalite are common. The presence of fayalite is marked by the replacement of grunerite and greenalite and transformation into magnetite and other sulphides. Iron-potassic alteration is common, creating schistosity in biotite units.

36 km9 9 18 27

Salobo(Cu, Au)

Ig. Bahia Mine (Au)

Ig. Azul Mine Azu(Mn)

Iron MineCarajás Parauapebas

Serra Pelada(Au, Pt, Pd)

EFCSL

Curionópolis

Depósito Cristalino

(Cu, Au)

Canaã dos Carajás

Vermelho(Ni)

118

(Cu, Au) Sossego

(Cu, Au)Serra Sul

(Fe)

Basement > 2.25 Ga

Mafic-Ultramafic Layered Complex

Andorinhas Supergroup

Pium ComplexXingu Complex

Uatumã Group - 1.8 Ga Tocantins Gropu -1.0 Itacauinas Supergroup ~ 2.7 Ga

Buritirama Group

Grão Pará Group

Ig. Bahia Group

Aquiri GroupIg. Pojuca GroupIg. Salobo GroupPlaquê Suite

Granitic Bodies

Dioritic Bodies / Gabbroic Complex

Mafic-Ultramafic Layered Complec

Tocantins Group

Grão Pará Group Xingu Complex

Grão Pará Group - BIF Pium Complex

Igarape Pojuca Group Aquiri Group Andorinhas Supergroup Uatumã Group

Aguas Claraas Formation

Plaquê Suite Igarape Bahia Group

Igarape Salobo Group

Depósito Cristalino

(Cu, Au)

December 2019 25

The southeast portion of the deposit hosts hastingsite, replaced partially by actinolite, grunerite and sulphide minerals. Fluorite, apatite, graphite and uranium oxides are associated with this assemblage, Fe-silicate minerals and alteration products of fayalite.

Garnet-Grunerite Schist (DGRX)

These are massive rocks with local development of schistosity. The rocks with significant almandine and grunerite content have isotropic texture or very few schistosity structures, with nematoblastic and granoblastic texture. The main mineralogical composition consists of almandine and cummingtonite-grunerite, with magnetite, hematite, ilmenite, biotite, quartz, chlorite, tourmaline and subordinate allanite. Fluorite and uraninite generally occur in veinlets related to stilpnomelane, calcite and grunerite.

Biotite Schist (BDX)

This unit is the most common lithology at Salobo and consists of medium to coarse-grained material with anastomosed foliation. The mineral assembly is characterized by biotite (responsible for the foliation observed within the rocks), garnet, quartz, magnetite and chlorite. The assemblage with garnet, magnetite, grunerite and biotite is partially replaced by a second generation of biotite and magnetite with chlorite, K feldspar, quartz, hematite and sulphides. Tourmaline, apatite, allanite, graphite and fluorite generally occur throughout this unit.

Feldspar-Chlorite Mylonite (ML)

The feldspar-chlorite-quartz mylonite is characterized by mylonitic foliation, produced by the orientation of rims of chloritized deformed biotite, hastingsite, elongated quartz and saussuritized plagioclase (K-feldspar, epidote and muscovite alteration). Porphyroblastic garnet is partially or totally replaced by chlorite and epidote. Allanite and apatite generally occur throughout this lithology.

Metavolcanic Basic (MTB)

This group of massive coarse-grained rocks is characterized by Fe-hastingsite and/or hornblende and plagioclase with chlorite alteration. It occurs irregularly in the system, but is concordant with other lithotypes in abrupt contacts, probably hydrothermally altered intrusive basic relicts within the package of volcanic rocks.

Quartz Mylonites (QML)

Quartz mylonite is grey or white in colour, passing through green to red. Where present, Fe-oxides are medium to fine grained, foliated and composed predominantly of quartz, muscovite, sericite, sillimanite and chlorite. Accessories,

December 2019 26

such as biotite, feldspar, magnetite, almandine, tourmaline, zircon and allanite are common. It is possible to differentiate: (a) red quartz-feldspathic rocks formed by K-feldspar and quartz and which may be a product of shearing between the gneissic basement and the supracrustal rocks; and (b) chlorite schists, mainly composed of chlorite and quartz, that represent intense hydrothermal alteration. This unit is found near the southern border of the deposits, close to important brittle shear zones, which may be interpreted as conduits for hydrothermal fluids.

Old Salobo Granite (GR)

The Old Salobo Granite occurs as a stockwork of approximately 2,573 ±2 Ma. The rocks appear colorless-pink to grey, coarse grained and with mylonitization in some areas. The main mineralogy is composed of K-feldspar (orthoclase-microcline), oligoclase, quartz, augite, hornblende, chlorite and, rarely, magnetite. There is no evidence of contact metamorphism with the host rocks. The mylonitic aspects that appear both in granite and host rocks are likely to have formed during the deformation phase.

Young Salobo Granite (GR)

The Young Salobo Granite occurs as small northwest-trending sills, hosted by the supracrustal sequence and by the gneisses of basement. It corresponds to the youngest granitic intrusion detected by drilling in the Salobo area. In some porphyritic portions, the matrix is aphanitic, containing a porphyry of red albite (Fe-oxide in micro-fractures) and chlorite pseudomorphed by biotite. This mineral assemblage is composed of fine to medium grained, equigranular, hypidiomorphic grains of albite/oligoclase, orthoclase, quartz, chlorite, with minor epidote, zircon, fluorite, magnetite, chalcopyrite and pyrite. Deformation was not observed, and the structure is isotropic. Age dating indicates an age of 1,880 ±80 Ma.

Diabase (DB)

Diabase is located in southeast of the deposit, striking at approximately N70°E, while in the northwest of the deposit striking near to N20°W. The predominant minerals comprising the rock type are augite, plagioclase, magnetite, ilmenite and quartz. The fine-grained diabase has an age of 553 ±32 Ma, while the more granular margins are dated at 561 ±16 Ma. This unit represent the last magmatic event of the area. The dykes are set within shear/fault lateral geometries to (N70°E) and frontal geometries (N20°W), probably developed before the intrusions, in a compressional regime modified by an extensive regime.

Rhyolite (RIO)

Rhyolite dykes are grey-reddish in colour, porphyritic in texture within an aphanitic matrix. The majority are composed of K-feldspars, plagioclase, quartz, amphibole

December 2019 27

in a matrix cut by quartz veinlets. In drill holes the occurrence is rare or an ultimate phase.

Figure 4 - Major Lithological Units – Plan View

Figure 5 - Major Lithological Units – Vertical Section View

December 2019 28

Figure 6 – Core Photos of Major Lithological Units

7.3. Tectonic Setting As depicted in Figure 7, the Salobo deposit is situated within the Cinzento strike-slip system which has been described as a set of Archean alignments that forms the Salobo transpressive duplex (or Salobo sidewall rip-out). This system post-dates the formation of the Itacaiúnas shear zone and was developed under ductile–brittle to brittle conditions.

Figure 7 - Tectonic Setting of the Carajás Region

Note: Figure from SMSA/CVRD, 2003a. Salobo 3 Alfa is the Salobo deposit.

N

50 km

Salobo

3 Alfa

December 2019 29

The tectonic evolution of the Salobo area includes sinistral, transpressive, ductile deformation that developed under upper-amphibolite-facies conditions, followed by sinistral, transtensive, ductile–brittle-to-brittle shear deformation. Shear zones are characterized by a mylonitic, penetrative foliation that generates a compositional banding. Where deformation is more intense, S-C foliations are parallel, and a lenticular pattern develops. The ductile deformation along the Itacaiúnas shear zone, which has affected the basement rocks and rocks of the Salobo Group, produced widespread, subvertical, northwest–southeast schistosity, which affects all lithologies in the deposit, except the Young Salobo Granite and the diabase dykes. The transtensive deformation along the Cinzento strike-slip fault system reactivated old structures and formed a subparallel ductile–brittle shear zone in the northern part of the deposit and a brittle shear zone in the south. Brittle–ductile shear zone deformation has resulted in lenticular-shaped ore shoots that characteristically show close associations between copper mineralization and magnetite content.

7.4. Metamorphism Two phases of metamorphism have been recognized in the Salobo area:

• Initial phase: associated with progressive amphibolite-facies metamorphism developed under ductile conditions of high temperature (650˚C), low pressure (2–3 kbar), and oxygen fugacities of -20 and -18. This caused partial substitution of chalcopyrite by bornite and chalcocite, accompanied by intense K-metasomatism

• Retrograde phase: developed under greenschist facies, with an average temperature of 340˚C; characterized by intense chloritization and partial substitution of bornite by chalcocite.

7.5. Alteration The Salobo hydrothermal system has a core of massive magnetite that is surrounded by less intensely altered rocks. Within the massive magnetite body there are small veins and irregular masses of secondary biotite. Garnet is completely replaced by magnetite, forming pseudomorphs. Away from the massive magnetite, the magnetite content gradually diminishes, giving way to biotite–garnet schist and/or garnet–grunerite schist. Alkali-metasomatism of the amphibolite facies rocks is expressed by weak sodium with intense, superimposed potassium alteration (≤4.6 wt% of K2O). K-feldspar, biotite and oligoclase are the main alteration minerals. A significant increase in the FeO content (≤35 wt%) accompanied the potassium alteration in amphibolite, and

December 2019 30

was marked by the replacement of calcium-amphibole (mostly magnesium-hornblende and hastingsite) by iron–magnesium amphibole (cummingtonite), and by formation of biotite and magnetite. The chemistry of the meta-greywackes at the deposit indicates that they also underwent significant iron and potassium alteration. Alteration assemblages are characterized by almandine, garnet, biotite and grunerite, subordinate tourmaline and minor magnetite. The better-mineralized zones, located in the central part of the deposit, correspond to the most altered areas.

7.6. Mineralization The Salobo deposit extends over an area of approximately 4 km along strike (west–northwest), is 100–600 m wide, and has been recognized to depths of 750 m below the surface. The sulphide mineralization typically consists of assemblages of magnetite–chalcopyrite–bornite and magnetite–bornite–chalcocite. Accessory minerals include hematite, molybdenite, ilmenite, uraninite, graphite, digenite, covellite, and sulphosalts. The mineral assemblages can be found in a number of styles: forming disseminations, stringers, stockworks, massive accumulations, filling fractures, or in veins associated with local concentrations of magnetite and/or garnet filling the cleavages of amphiboles and platy minerals, and remobilized in shear zones (Figure 8). There is a positive relationship between copper minerals and magnetite. Copper content is typically >0.8% in XMT and BIF, whereas in gneisses and schists it is <0.8%. A positive correlation between copper content and uranium contents has also been established. Chalcopyrite, bornite, and chalcocite occur interstitially to silicate minerals. These sulphide minerals are commonly found filling cleavage planes of biotite and grunerite. Hematite is rare, but in places it can reach as much as 4% by volume. It exhibits tabular textures (specularite), with infilling bornite, and partial replacement by magnetite. Native gold occurs as grains smaller than 10 μm in cobaltite, safflorite, magnetite and copper sulphides, or interstitial to magnetite and chalcopyrite grains. Native gold grains contain up to 10 wt% Cu, with subordinate silver, arsenic, and iron. Molybdenite occurs interstitial to magnetite and shows cleavage planes filled with chalcopyrite and bornite. In mylonitic samples, molybdenite forms kinked stringers. Magnetite occurs mainly as idiomorphic to sub-idiomorphic grains, interstitial to silicate minerals or in fractures, or forms bands in mylonitic rocks. The gangue minerals are almandine garnet, grunerite, and tourmaline, reflecting the intense iron-metasomatism. Minor amounts of fayalite and hastingsite are

December 2019 31

pseudomorphed by grunerite and magnetite. Tourmaline, with a dominant schörlitic (black-tourmaline) composition, occurs as idiomorphic crystals preferentially oriented parallel to mylonitic foliation, in association with biotite, garnet and grunerite. Ilmenite, uraninite, allanite, fluorite and apatite occur as accessory minerals. Biotite sub-idiomorphic crystals, commonly kinked, are associated with potassic alteration, and spatially related to the copper–gold mineralization. Uraninite and zircon inclusions may be locally abundant in biotite. Quartz is associated with biotite in ore-grade samples and forms concordant veins within the host rocks. Textural relationships indicate that mineralization was developed firstly as an oxide stage, with a second, subsequent, sulphide stage.

Figure 8 - Copper Mineralization Styles at Salobo

7.7. Comments on Section 7 In the opinion of the QPs, the knowledge of the deposit settings, lithologies, mineralization style and setting, ore controls, and structural and alteration controls on mineralization is sufficient to support Mineral Resource and Mineral Reserve estimation.

8. Deposit Types There appear to be two classes of copper–gold deposits in the Carajás region. The first group includes Cu–Au–(W–Bi–Sn) deposits which contain quartz veins, and may or may

December 2019 32

not have associated iron oxides and are genetically related to the cooling of Palaeoproterozoic (ca. 1.88 Ga) granites. The second group includes iron oxide Cu–Au (±U–rare earth elements) deposits (e.g., Salobo, Sossego, Cristalino, 118 and Igarapé Bahia) that may be related to more alkaline rocks, including the ca. 2.57 Ga alkaline complexes of the Carajás belt (e.g., Estrela Complex, Old Salobo Granite) and the base metal mineralization-associated 1.88 Ga intrusives. The second group of deposits are commonly referred to as iron oxide copper gold deposits (IOCG). Global examples of IOCG deposits include Olympic Dam in Australia, Candelaria–Punta del Cobre in Chile, and Sossego in Brazil.

9. Exploration The discovery of the Salobo copper deposit occurred during a systematic program of geochemical, geophysical and geological exploration in the Carajás region, initiated by CVRD/Docegeo in 1974. Since then, the area has been the subject of exploration and development activities and a considerable information database has developed as result of both exploration and mining activities. Table 3 summarizes the exploration activities from 1978 to 2003.

Table 3 - Exploration Summary 1978–2003

Activity Details (unit) Docegeo CVRD/GICOR SML SMSA CVRD/SMSA

Total 1978 1986 1993 1997 2002–2003

Surveying Area (ha) - - - 3,091 - 3,091

Lines (km) 258.2 21.3 - 52.3 - 52.3

Geochemistry Lines (km) 3,230 138 - - - 3,368

Samples 3,433 2,616 - - - 6,049

Geophysics

I.P. (km) 26.3 165.9 - - 180.6 180.6

Magnetometry (km) 76 171.6 - 43.5 212.7 256.2

Scintillometry (km) 52.3 - - - - 52.3

TEM (loops) 750 750

Gamma-spectrometry (km) - - - 26.3 214.5 240.8

Shafts Amount 18 23 6 - - 47

Length (m) 54 377.2 93.5 524.7

Adits Amount 2 1 - - - 3

Length (m) 450 950 - - - 1400

Mapping Lines (km) 427.6 221.3 1.8 43.5 - 694.2

No exploration occurred at Salobo between 2003 and 2011. In 2012, a regional airborne gravity survey was completed. The survey identified a potential continuation of the Salobo orebody at depth. In 2017, a deep drilling campaign was initiated exploring the deep extension and potential for underground mining. At the time of the Report five holes had been completed with assays received on all except one and a sixth hole was in progress. All holes encountered mineralization below the current Mineral resource pit.

9.1. Geological Mapping Geological mapping at different scales was conducted over the Salobo deposit during the initial campaigns, usually following survey traverses. However, due to the fact that nearly 80% of the rocks in the Carajás regions are poorly exposed, most direct observations

December 2019 33

were made along access roads for drill sites and were complemented with additional information such as interpretation of air-photo images, geophysical and geochemical maps, and correlation on surface of core logging data. Current pit mapping is conducted twice a month. A geologist loads the long-term geologic map over the updated topographic map and establishes the actual position of the geological contacts in such points where access is possible.

9.2. Airborne Gravity Survey A regional airborne gravity gradiometer survey was completed in 2012 over a portion of the Carajás Region, including the Salobo Operations area. It was designed to explore for new shallow copper–gold targets. The survey flight is typically at an altitude of 80 m or greater with a line spacing dependent on the target of investigation. The 2012 survey was flown on 100 and 200 m spaced lines. The aero-gravity gradiometer method measures acceleration of gravity and gradients of the acceleration of gravity respectively. It is not a precise tool and is not robust below 500 m depth. There are a number of viable interpretations of the gravity data, one of which is that there could be a vertical extension of the Salobo mineralization below the current planned open pit, as the geological data have already indicated. In 2017, a deep drilling campaign was initiated exploring this potential orebody extension. At the time of this report, five holes had been completed with assays pending on one and a sixth hole was in process. All holes encountered mineralization below the current Mineral resource pit. Figure 9 shows the gravity gradiometer survey results in relation to the magnetic survey in the Salobo Operations area. Figure 10 shows the gravity model, which has the 2015 block model outline at a 0.5% Cu cut-off, superimposed.

December 2019 34

Figure 9 - Coincident Magnetic and Gravimetric Anomalies

Note: Figure courtesy Vale, 2015. White outline is the current extend of the Salobo mineralization, projected to surface

Figure 10 - 3D Gravity Inversion with Current >0.5% Cu Block Model Outline

December 2019 35

9.3. Comments on Section 9 In the opinion of the QPs, the Salobo deposit has been explored using appropriate techniques. Geophysical surveys completed by the Exploration department at the Salobo Operations have identified a significant gravity anomaly below the current Salobo open pit. Deep drilling has confirmed mineralization below the currently modeled Mineral resource pit. Further work is required to determine what the exploration potential at depth and help to identify additional Mineral Resources for supporting future potential operations and projects expansions.

10. Drilling Diamond drill hole core is the majority sample type for geological modelling and mineral resource estimation at Salobo. Blast holes have been drilled since 2009 but are used only for grade control and short-term planning. Core drilling commenced in 1978 and was conducted through to 2003 in five different drilling campaigns, for a total of 420 holes completed for exploration purposes, and an additional 15 drill holes for geotechnical purposes (Table 4). Most drill holes were vertical or oriented to the south–southwest, the latter with dips usually ranging from 60° to 70°. However, one campaign included holes with a north–northwest orientation and similar dips. Various holes were also drilled from an adit. No exploration core drilling occurred between 2003 and 2016. In 2010, two infill holes were completed and in 2017, infill and deep drill programs were initiated at the mine. The following table summarizes the drilling campaigns completed at the Salobo Operations and Figure 11 is plan-view map of the drill hole traces.

December 2019 36

Table 4 - Drill Hole Summary Table

Campaign/Period Purpose Drill Hole ID Total Meterage Drilled

(m) Percentage of total drilling

(%)

1978 Exploration SAL-2ALF-FD001 to SAL-

3ALF-FD 065 29,275 15%

1986 Exploration SAL-SALF-FD066 to SAL-

3ALF-FD 125 9,051 5%


3ALF-FD 189 14,585 8%


3ALF-FD 277 25,491 13%


3ALF-FD 420 69,908 36%

2010 Infill SAL-3ALF-FD421 to SAL-

3ALF-FD 422 361 0.2%

2017 Infill S3A-FD00423 to S3A-

FD00464 13,264 7%


FD00505 12,674 7%


FD00533 10,159 5%

Total exploration 184,768 96%

1997 Geotechnical SAL-3ALF-FG001 to SAL-

3ALF-FG 007 3,847 2%

2003 Geotechnical SAL-3ALF-FG008 to SAL-

3ALF-FG 0134 4,194 2%

Total geotechnical 8,042 4%

Grand Total 192,810

December 2019 37

Figure 11 - Plan View – Drill Hole Traces

Figure 12 is a section showing three holes from the infill drilling program and three from the deep program with accompanying assay summaries. Holes drilled in prior campaigns are not displayed. The approximate location of the section is shown in Figure 11. The infill holes hit significant mineralized intervals and the deep drill holes encountered mineralization below the currently modeled Reserve pit.

December 2019 38

Figure 12 – Infill & Deep Drilling, Section 1200E

10.1. Drill Methods

All boreholes drilled in support of Mineral Resources were planned and laid out according to the Vale Salobo Guidelines and Reporting Standards. Compliance with these guidelines is checked on a quarterly basis to ensure guidelines and procedures are being followed. Surface drilling was typically initiated with HQ diameter (63.5 mm) core and reduced to NQ diameter (47.6 mm). The minimum diameters were BX (36.6 mm) and BQ (36.5 mm). The underground drilling utilized BX diameter rods. Drill hole collar locations are determined with a total station survey instrument. Collar verification is completed by plotting drill hole locations on plan and in cross section and comparing with the topographic surface. Down hole surveys are verified against the original survey data and on cross section plots.

10.2. Core Reception, Handling and Storage The diamond drill core is collected and placed in wooden boxes. Core is delivered by the drilling contractor to the core logging / storage area where geological and geotechnical logging is carried out. At the core logging facility, core recovery and physical properties are measured and recorded. Geologic logs are prepared, and sample intervals are

December 2019 39

marked. Sample intervals average 1 m in mineralization and 2 or 4 m in barren zones. Sample lengths vary from these standards to honor significant geologic boundaries. Before logging is completed pictures are taken of all drill core. After logging, the mineralized core is split in half using an electric saw, with one half being retained for further studies and audit purposes, whereas the other half is submitted for sample preparation and analysis. The position where the core is to be split is marked by Vale geologists. Facilities for drill core storage consist of warehouse facilities special for this purpose and are located at the project site. Drill cores are stored in wooden trays and labelled with metallic plates. Pulps are stored in paper envelopes grouped in plastic bags, while the coarse rejects are stored in plastic bags. Both are organized in properly identifies boxes. Since 2017, the mine site core facility has been used. A new core shed is being built in the town of Parauapebas which the mine will transition to when completed.

10.3. Geological logging No written details are available on the logging procedures for the drilling campaigns prior to 1997. During drilling campaigns from 1997 onward, core was collected in 1 m wooden boxes, and photographed in sets of two boxes each after transportation to the core shack. Logging was completed after sampling, and consisted of describing each individual lithologic package, as well as mineralogical variations within each one, the textures and structures, the ore minerals (including a visual assessment of volume percentage), the presence of deleterious minerals (mainly fluorite), the visible structures and the foliation angle with respect to the core axis. As part of the logging procedure, magnetic susceptibility was measured using Scintrex K2 and KT5C kappameters, with readings every 20 cm. This information was recorded in paper format (De Souza and Vieira, 1998). In some of the early campaigns, uranium, thorium and potassium were directly determined in core using an Exploratium GR-320 gamma-spectrometer; however, this method was soon discarded. During the 2002–2003 campaign, uranium was chemically determined on 2 m intervals, which allowed this deleterious element to be modelled during Mineral Resource estimation. Geotechnical logging of drill core was conducted by geologists following the guidance of geotechnical engineers. Logging included simple descriptions of the weathered zones and the weathering and fracturing degrees of the mineralized schists, as well as visual determination of the rock-quality designation (RQD) and rock resistance, and descriptions of the fracture types. Point-load tests (PLT) were also conducted every 20 m.

December 2019 40

10.4. Recovery Micon (2013) noted that core recoveries of 80% in weathered rock and 90% in fresh rock were achieved by the drilling companies during the campaigns. The average core recovery of the 2002 campaign (drill holes SAL-3ALF-FD 278 to 410) was 97.6%. AmecFW (2016) reviewed the recovery data for holes where the information was recorded and supported Micon’s assessment of overall good recoveries. Core recovery for the infill and deep drilling since 2017 has averaged 99%.

10.5. Collar Surveys During the 1997 campaign, drill-hole collars were placed and resurveyed after completion using WILD T1 stations. During the 2002–2003 campaign, drill sites were placed and collar coordinates measured using total station equipment (before and after hole completion). The survey team also oriented the drill rigs, and provided proper initial alignment and inclination to the drilling rods. Collar verification was completed by plotting drill hole locations on plan and in cross-section and comparing with the topographic surface. Current collar surveying is conducted by company surveyors using high-precision, differential GPS equipment.

10.6. Downhole Surveys No written details on the down-hole survey procedure in place prior to the 1997 campaign were available to AmecFW for their 2016 technical report. During the 1997 campaign, down-hole survey readings taken on average every 3 m were conducted using the Reflex DDI (dip and direction pointer) and Maxibor units, to prevent errors in azimuth readings due to the influence of magnetite in the host rocks. During the 2002–2003 campaign, down-hole survey measurements were conducted every 3 m using Reflex Maxibor and gyroscopic instruments. Since 2017, the infill and deep drilling have utilized the Reflex Gyro tool for downhole surveys and switched in 2019 to Reflex Gyro Sprint which is also used for the deep drilling downhole surveys.

December 2019 41

10.7. Specific Gravity Determination During the 1997 campaign, bulk density determinations were made with the water-displacement method. Tests were conducted on 20 cm to 40 cm long saprolite and bedrock core samples within intervals of approximately 10 m length. Wet and dry bulk densities were determined on saprolite samples, which were weighed in air prior to and after drying (respectively), then coated with a thin plastic film, and submerged in water in a PVC recipient with a discharge opening. The sample volume was determined by measuring the water displaced through the discharge into a graduated cylinder. Core samples were assumed to be dried, so only dry density was determined. The bulk density (D) was determined as D = P/V, where P is the dry (or wet) weight, and V is the volume of displaced water. During the 2002–2003 campaign, the specific gravity (SG) was determined on representative fragments from all sampling intervals using a standard procedure. Hard-rock samples were cleaned and dried in air, and then weighed in air and in water. Saprolite samples were dried using an oven, then coated with paraffin prior to submerging them in water. Since 2017, the infill and deep drilling programs are utilizing the same methods as the 2003 drilling campaign for SG measurements. SG was then estimated as follows: SG = xA / (xA – xW) Where: xA = weight of core in air xW = weight of core completely submerged in water At Salobo, SG was measured on approximately 84,000 samples collected across the entire deposit. Values for weathered waste rock and unweathered bedrock were categorized separately due to differences in permeability and porosity caused by weathering.

10.8. Comments on Section 10 The QPs have reviewed the drill methods, core reception, handling and storage, geological logging, recovery, collar surveys, downhole surveys and specific gravity determination as well as the work done by independent third-party consultants. The QPs have not identified any drilling, sampling or recovery factors that could materially impact the accuracy and reliability of the results in reviewing those programs. In the opinion of the QPs, the quantity, quality, results and interpretation of the exploration and infill drill

December 2019 42

programs during the 1997 and later campaigns are sufficient to support Mineral Resource and Mineral Reserve estimation.

11. Sample Preparation, Analyses and Security

11.1. Sampling Methods

11.1.1. Drill Core The core sampling procedure was similar during the various campaigns since 1997. Sample intervals averaged 1 m in mineralized zones, and between 2 m and 4 m in barren zones. Sample intervals are constrained to geologic boundaries, lithological / mineralogical changes, and faults and shear zones. Core is halved using diamond saws with one half bagged and submitted to the mine laboratory for analysis, and the remaining half retained as backup in the same original boxes.

11.1.2. Blast Holes Blastholes are drilled on a 5 m x 5 m or 5 m x 7 m grid with a hole diameter of 12¼ inches. All blastholes located in ore zones are sampled; however, as the blasthole reaches the barren zones, the proportion of sampled holes decreases (one in two or even less), and the grade-control geologist determines which waste blastholes are sampled to ensure mineralization matches the interpretation in the geological model. The sampling pattern depends on the shape of the cone. If it is well formed, then four channels are cut across the cone at 90° (Figure 13) using a small mattock, and the sample is collected using a jar from bottom to top of the inner channel wall. If the cone has been partially damaged, then three channels are cut; however, if it is seriously damaged then the cone is not sampled. The average sample weight is 2 kg. Tags are being inserted into the ore after blasting for tracking purposes to understand the differences between the samples estimated recovery and processing plant actual recovery. The tags are detected at specific points in the processing plant.

December 2019 43

Figure 13 - Blast-Hole Sampling Pattern


11.2. Sample Preparation

11.2.1. Exploration Sample preparation details prior to 2002 are unknown. During 2002 – 2003, sample preparation was conducted by Lakefield / GEOSOL laboratory at a local facility built at the Project site, and consisted of the steps detailed in Figure 14.

Channels in the blast-hole cone

December 2019 44

Figure 14 – 2002 – 2003 -Sample Preparation Flowchart

The infill and deep drilling programs that began in 2017 are using sample preparation procedures similar to the 2002 – 2003 campaign and are detailed in the flowchart shown in Figure 15. Sample preparation was being done at the Salobo mine laboratory but was switched to ALS, Vespasiano, Minas Gerais, Brazil in December 2018 in order to reduce the backlog of pending samples.

Every 20th Sample has the

following QA/QC sample

100 g Lak-Geosol duplicate

100 g Arbitro Lab duplicate

40 g GAMIK Au duplicate

20 g GAMIK duplicate

40 g GAMIK pulp reject

100 g to

SATMAGAN

Pulp Reject

to store

Pulverized 95% < 0.105 mm

Homogenized

Rotary Splitter

Every 20th Sample for

Duplicate Analysis

100 g to ACME

C, S, U, Mo & F

100 g to Lak_Geosol

Cu, Au & Ag

Core Sample

Dried at 105°C

Crushed to 95% < 4 mm

Homogenized

Jones Riffle Splitter

Coarse Reject

December 2019 45

Figure 15 – 2017 - 2019 -Sample Preparation Flowchart

11.2.2. Grade Control Blast-hole samples are prepared and assayed at the Salobo Operations laboratory which has separate areas for the preparation of concentrate, tailings and blast-hole samples to avoid contamination. The preparation laboratory is well organized, and has modern equipment including ESSA jaw crushers, rotary splitters, puck-and-bowl pulverizers and Mettler-Toledo precision scales. A special, separated, scale room is used only for gold assays. The dust-extraction system is in place to reduce the chances of sample contamination. The preparation procedure implemented for blast-hole samples is as follows:

• Drying in an electric oven at 105°C

• Jaw-crushing to >95% passing -3 mm size; granulometric tests are carried to check particle size on one in 20 samples

• Homogenization and splitting using rotary splitters to obtain 500 g splits

• Pulverization using puck-and-bowl pulverizers to >95% passing 0.105 mm; granulometric and mass-loss checks are carried out on one in 20 samples on 100 g subsamples that are later discarded

• The pulverized material is bagged and submitted for chemical assay.

Core Samples

Dried at 105°C

Homogenized

Rotary Spitter

Homogenized

Rotary Splitter

Crushed to 95%

<25mm

Pulp Reject stored150 g pulps for

analysis & QA/QC

Pulverized 95%

passing <0.106 mm

Crushed to 95% <

3.35mm

Coarse Reject stored

December 2019 46

11.3. Sample Analysis

11.3.1. Exploration During the 1978 campaign, samples were assayed at the Docegeo laboratory in Belém, Pará, and at the SUTEC laboratory in Santa Luzia, Minas Gerais. Copper was assayed on 0.5 g aliquots by multi-acid digestion and atomic absorption spectrometry (AAS). Iron, molybdenum, and silver were also determined using this method. Gold was assayed by aqua regia leaching, with solvent extraction (MIBX) and AAS determination. During the 1986 campaign, CVRD assayed samples at the Docegeo laboratory in Belém and at the pilot plant laboratory on the mine site, using the same analytical methods as in the previous campaign. During the 1993 campaign, SML used the Mineração Morro Velho (MMV) laboratory. Copper was assayed with multi-acid digestion and AAS reading on 0.5 g aliquots (0.002% detection limit), and gold was determined using the fire-assay (FA) method with gravimetric finish on 100 g aliquots (0.05 g/t detection limit). In addition, samples were assayed for sulphur and carbon by LECO, and fluorine by alkaline fusion with sodium carbonate and potassium nitrate, followed by ion-selective electrode determination. SMSA used the same analytical procedures during the 1997 campaign. During the 2002–2003 campaign, Lakefield GEOSOL was used for the routine analysis of copper, gold, and silver, while Acme analyzed for molybdenum, uranium, fluorine, sulphur, and carbon. Chemical analysis was by atomic absorption spectrometry (AAS) for copper, silver and fluorine (on a 0.5 g aliquots and multi-acid digestion), while gold was assayed by FA with AAS finish on 20 g aliquots. Sample rejects are currently kept stored at the mine core shack. In the early stages of the exploration program platinum, palladium, nickel, molybdenum and uranium were also analyzed; however, these elements were later excluded from the analytical package. The infill and deep drilling campaigns which began in 2017, are using ALS, Lima, Peru as the primary lab utilizing the following analytical methods:

• Copper analysis by four-acid digestion and atomic absorption reading (Cu-AA62)

• Gold by FA of 50g aliquot, two-step digestion with nitric and hydrochloric acids and reading by atomic absorption (Au-AA24)

• Multi-element (including main elements and traces, in addition to copper, sulfur and uranium) with digestion by four acids and readings by ICP-MS or ICP-AES (ME-MS61).

December 2019 47

The secondary laboratory used is SGS-Geosol in Vespasiano, Minas Gerais, Brazil utilizing the following analytical methods:

• Copper analysis by multi-acid digestion and atomic absorption reading (AAS41B)

• Soluble copper by acetic acid digestion and atomic absorption reading (AAS51C)

• Gold by FA of 50g aliquot, digestion with aqua regia and reading by atomic absorption (FAA505)

• Multi-element (including main elements and traces, in addition to copper, sulfur and uranium) with digestion by four acids and readings ICP-AES or ICP-MS (ICM40B)

• Carbon and sulfur by LECO (CSA17V)

• Fluorine by ion-specific electrode (ISE), (ISE03A).

11.3.2. Grade Control Blast-hole samples are assayed at the Salobo Operations analytical laboratory for copper, gold, silver, iron, carbon, sulphur, fluorine, chorine and soluble copper. Table 5 details the various analytical methods used for each of the elements and their detection limits.

Table 5 - Blast-Hole Sample Analysis

Element Aliquot (g) Method Detection Limit Quantification Limit

Cu (%) 0.25 ARD-AAS 0.003 0.006 Au (g/t) 30 FA-AAS 0.004 0.006 Ag /g/t) 10 MAD-AAS 0.132 0.202 Fe (%) 0.25 ARD-AAS 0.018 0.030 C (%) 0.25 LECO 0.002 0.003 S (%) 0.25 LECO 0.001 0.003

F (ppm) 0.3 AF-ISE 94 139 Cl (ppm) 1.0 SAL-SNT NA NA

CuSol (%) 1.0 AcAL-AAS 0.001 0.003

Note: ARD: aqua-regia digestion; AAS: atomic absorption spectrometry; FA: fire assay; MAD: multi-acid digestion; AcAL: acetic acid leach; AF-ISE: boric-acid/sodium-carbonate fusion and ion-selective electrode determination; SAL-SNT: sulphuric acid leach and silver-nitrate titration; NA: not available; CuSol – acid soluble copper.

Precision scales and assay instruments are linked to a laboratory information management system (LIMS) to ensure the assay data are digitally transferred into the mine database. The LIMS is programmed to determine when readings comply with the required quality-control thresholds. Turnaround time is usually less than 24 hours for most elements, and four to five days for fluorine and chlorine. Coarse and pulp rejects are stored for 3 months, after which they are discarded. Assay batches are usually organized in 25 samples, not including the internal control samples. The lab’s quality control (QC) protocol includes the insertion of one reference material, one reactive blank (consisting of pure solution or flux in the case of FA), one coarse duplicate, and one pulp duplicate per batch.

December 2019 48

11.4. Quality Assurance and Quality Control The quality control (QC) programs at the Salobo Operations varied considerably over time, depending on the primary analytical laboratory used for assaying.

• 1986 – A total of 402 samples were resubmitted to alternative laboratories for external checks with GEOSOL acting as secondary laboratory for the Docegeo laboratory for copper and gold assays, the pilot plant laboratory as secondary laboratory for Docegeo on copper assays and Docegeo as secondary laboratory for the pilot plant laboratory for gold assays.

o Results on copper assays indicated good correlation between the three

laboratories; however, poor correlation was obtained between GEOSOL and Docegeo on the gold assays.

• 1993 - The QC program included external checks of 5% of the samples at the Nomos laboratory (for Cu) and at Fazenda Brasileira (for Au), using the FA method. In total, copper checks were conducted on 664 samples, and gold checks on 2,168 samples. For both elements, the correlation between laboratories was assessed as good.

• 1997 - SMSA implemented a QC program consisting of the insertion of 574 coarse duplicates and 14 reference materials, and the submission of 750 check samples to the Label laboratory for external checks.

• 2002 - Due to the lack of appropriate QC results for the drilling campaigns prior to 2002, a re-assay campaign was initiated to validate the available analytical data, thus a total of 51,768 of the original 75,577 samples drilled prior to 2002 were re-assayed to corroborate the original results.

o Vale concluded that the external assay check review revealed bias for copper and gold assay results obtained by Nomos and Gamik laboratories. Based on the results obtained, Vale applied an adjustment factor to original sample grades (Table 6 and Table 7).

Table 6 - Adjustment for Copper Assays for pre-2002 Drilling Programs

Holes Number Outliers Interval Regression Correlation Coefficient

Docegeo D-001 to

065 10,833 126 1% >0.01 Cu adj = (1.029 * Cu) + 0.007 98%

CVRD D-066 to

125 3,609 113 3% >0.02 Cu adj = (1.068 * Cu) – 0.02 98%

SML D-126 to

189 400 46 12% >0.01 Cu adj = (0.98 * Cu) + 0.023 97%

SMSA D-190 to

277 12,453 489 4% >0.01 Cu adj = (1.014 * Cu) – 0.005 97%

Lakefield Geosol

D-278 to 410

1,440 33 2% >0.01 Cu adj2 = (0.997 * Cu adj) –

0.003 99%

December 2019 49

Table 7 - Adjustment for Gold Assays for pre-2002 Drilling Programs

Holes Number Outliers Interval Regression Correlation Coefficient

Docegeo + CVRD + SML D-001 to

189 26,760 522 2% >0.01 Au adj = (1.027 – Au) + 0.008 0.94

SMSA D-190 to

277 11,519 257 2% >0.01 Au adj = (1.018 – Au) + 0.005 0.85

• 2002-2003 - In-house Standard Reference Material (SRMs) samples used during the 2002–2003 campaign (a total of nine) were derived from both the sulphide and oxide mineralization and incorporate a significant spread in the copper and gold grades. The recommended values for SRMs were established from a set of analytical results provided by three laboratories (the former Bondar Clegg laboratory, Gamik and Lakefield / GEOSOL). Each laboratory analyzed 10 aliquots of each SRM.

o Two internal SRM samples were also prepared; however, they became

available only at the end of the drilling program. As a result, a total of 1,500 samples from the 2002–2003 drilling program were selected for re-assaying in order to validate the 2002–2003 assay data. A total of 76 samples of two internal, project-derived SRMs were randomly inserted in the batch (5% frequency).

• 2016 - AmecFW reviewed the QC data reported by CVRD (2003) and concluded that copper and gold check assays did not reveal significant biases, and that precision was within acceptable limits. Bongarcon (2003) also reviewed the 2002–2003 QC data, and concluded similarly that the special lot assays validated the 2002–2003 data for use in Mineral Resource estimation.

• 2017 – 2019 – With the recommencement of drilling in 2017, a QA/QC program was implemented aimed at ensuring the quality of physical preparation and chemical analyzes of samples and density tests. The program consists of the blind insertion of blanks, crushed duplicates, pulp duplicates, batch pulp duplicates, pulp duplicates to a secondary lab and certified standards. Table 8 shows the insertion frequency of each sample type.

December 2019 50

Table 8 - QA/QC Sample Types and Insertion Frequency 2017 - 2019

Code Type Insertion Frequency

CBK Blanks 5.0%

CDP Crushed Duplicates 2.5%

PDS Pulp Duplicates 2.5%

PDD Batch Pulp Duplicates 2.5%

CAS Pulp Duplicates - Secondary Lab 5.0%

STD

MSSO-002

5.0%

MSSO-003

OREAS 502

OREAS 503

OREAS 152a

ITAK 819

TH-1

TH-2

11.4.1. Blanks Golder (2010) reviewed the result of preparation blank samples submitted with the regular sample stream from 1999 to 2005. Although generally acceptable, some of the blank samples showed anomalously high content of copper which could have been caused by either cross-contamination in the sample preparation stage or sample mix-ups. All blank exceedances were followed-up with the laboratory and re-assayed. For the infill and deep drilling campaigns which started in 2017, blanks were inserted into the sampling stream at a frequency of 5% (Table 8). Results from analysis of blanks are plotted against the acceptable limits to determine if there is an indication of contamination that would require the sample batch to be reanalyzed.

11.4.2. Duplicates Three types of pulp duplicates were generated during the various Salobo QC programs including:

• Pulp duplicates, as part of the internal QC program of Lakefield Geosol.

• Inter-laboratory duplicates (external assay checks), part of the Salobo QAQC program. The majority of the pulp duplicates were sent as a special batch to a secondary laboratory (GAMIK) at the end of the program.

• A set of 1,500 samples from 2002-2003 drilling program, along with inserted standard samples, were re-analyzed to corroborate the original assay results by the primary laboratory.

The duplicate sample precision results are summarized in Table 9. For copper, the results demonstrate acceptable precision with average half absolute relative difference (HARD)

December 2019 51

values between 3% - 7%. This value is consistent with precision observed at 83.4% where values greater than 20% are indicating moderate precision. For gold the results indicate marginal to poor precision with average HARD values greater than 10%. This is consistent with precision observed at 83.4% where values are greater than 30%, although this may be expected due to its mineralizing pattern (nugget or fines) and therefore the levels of precision observed may be considered as acceptable. Bias is measured in terms of averaged HARD values. The results indicate no obvious bias for copper while for gold a slightly negative bias was detected.

Table 9 - Summary Duplicate Analysis

Element Laboratories Type #

Samples Avg HARD

(%) Avg HARD (%)

Avg Bias

Precision (at 83.4%)

Cu

Lakefield vs Lakefield Blind 515 7.37 0.06 0.01 28.7

Lakefield vs Gamik Blind 724 6.49 0.49 0.02 24.4

Lakefield vs Lakefield Non blind 429 3.16 -1.00 -0.01 21.2

Lakefield vs Gamik Non blind 667 6.32 1.28 0.00 23.7

Lakefield vs Gamik Special lot 1,502 6.47 -0.46 0.01 28.7

Au

Lakefield vs Lakefield Blind 515 16.98 5.97 -0.01 46.9 Lakefield vs Gamik Blind 537 15.92 -9.72 -0.05 41.3

Lakefield vs Lakefield Non blind 367 8.45 -0.68 0.00 32.2 Lakefield vs Gamik Non blind 355 21.46 -17.92 -0.09 46.2 Lakefield vs Gamik Special lot 1,417 13.88 -2.31 -0.04 42.8

For the infill and deep drilling campaigns which started in 2017, a number of different types of duplicates were inserted into the sampling stream (Table 8). These included crushed duplicates, pulp duplicates, batch pulp duplicates and secondary lab pulp duplicates and are used to monitor precision and accuracy of the sample preparation and analytical analyses. Results are regularly plotted against original results to determine if there is an indication of contamination that would require the sample batch to be reanalyzed.

11.4.3. Standard Reference Material For the infill and deep drilling campaigns that commenced in 2017, eight different standard reference materials (Standards) were used (Table 8) to monitor accuracy and precision of the analyses. Standards MSSO-002 and MSSO-003 were generated from materials from the Sossego mine. OREAS-152a, OREAS-503 and OREAS-502 are certified Standards purchased from Ore Research & Exploration (Australia). ITAK 819 is purchased from Copper Mining Company and TH-1 and TH-2 from the Thompson and Birchtree mines in Canada. The certified values for each Standard are shown in Table 10.

December 2019 52

Table 10 – 2017 – 2019 Standard Certified Values

95% Confidence

Limits

Standard Method / Element Certified Value 1SD Low High

MSSO-002 4 acid / Cu% 0.880 0.020 0.870 0.891

FA / Au ppm 0.079 0.002 0.076 0.082

MSSO-003 4 acid / Cu% 2.696 0.056 2.669 2.724

FA / Au ppm 0.285 0.007 0.279 0.291

OREAS 502

4 acid / Cu% 0.755 0.020 0.746 0.764

FA / Au ppm 0.491 0.020 0.482 0.499

OREAS 503

4 acid / Cu% 0.566 0.015 0.558 0.573

FA / Au ppm 0.687 0.024 0.676 0.697

OREAS 152a

4 acid / Cu% 0.385 0.009 0.379 0.391

FA / Au ppm 0.116 0.005 0.114 0.118

ITAK 819 4 acid / Cu% 1.080 0.028

FA / Au ppm 0.754 0.042

TH-1 Na peroxide fusion / Cu% 0.084 0.003 0.082 0.087

FA / Au ppm 0.054 0.008 0.050 0.059

TH-2 Na peroxide fusion / Cu% 0.157 0.003 0.154 0.159

FA / Au ppm 0.062 0.010 0.056 0.068

11.4.4. Quality Control for Blast-holes A QC program has been implemented to monitor blast-hole sampling quality. This program includes the insertion of 5% twin samples (obtained from repeating the sample process), 5% field duplicates (Jones splits of the same original sample that are assayed separately) and 5% SRMs.

11.5. Security During drill campaigns all drill core was brought from the drill sites, at the end of shift, to a dedicated logging and storage facility. All drill core is stored in wooden boxes with proper numbering to indicate the drill hole number and meterage. The core storage and logging facility is kept locked when unoccupied. Unshipped samples are also stored in a secure facility at the same location. Since August 2010, the evaluation of drilling and mine information has been uploaded to a Geovia Gems SQL database. This provides the geologists and mine engineers with a secure and more efficient access to information. In 2019, all short-range and long-range dataset is transferred to the GDMS database system, which is being used for drill core logging.

11.6. Comments on Section 11 In the opinion of the QPs, the 1997, 2002–2003 and current sampling, sample preparation, assay and density data are suitable to support Mineral Resource and Mineral Reserve estimation.

December 2019 53

12. Data Verification

12.1. Major Mining Studies Prior to commencement of production the following studies were completed:

• Bechtel, 1988: Prefeasibility study

• Minorco, 1998: Feasibility study

• Aker Kvaerner, 2001: Updated feasibility study

• Fluor JPS, 2004: Feasibility study.

12.2. External Audits and Reviews Vale and its predecessor companies have commissioned a number of audits and third-party reviews of block models and Mineral Resources and Mineral Reserve estimates and processes:

• MRDI, 1997: Mineral Resource estimate audit

• AMEC, 2002: Mineral Resource estimate audit

• AMEC, 2004: Mineral Resource estimate audit

• Pincock, Allen and Holt, 2007: Mineral reserve estimate audit

• Pincock, Allen and Holt, 2007: Due diligence audit

• Pincock, Allen and Holt, 2008: Mineral reserve estimate audit

• Snowden, 2009: Mineral Resource estimate audit

• Golder Associates, 2010: Mineral Resource and Mineral Reserve audit. Third-party reviews of data in support of a technical report prepared for Wheaton was completed in 2013 and 2016:

• Micon Consultants, 2013: Review of Mineral Resources and Mineral Reserves in support of a technical report filing for Wheaton.

• AmecFW, 2016: Review of Mineral Resources and Mineral Reserves in support of a technical report filing for Wheaton.

As part of AmecFW work preparing their 2016 report, they undertook various data verification procedures including the following:

• Review of drill-hole folders: Vale keeps ordered folders for each drill hole in the mine office. AmecFW reviewed 14 folders, corresponding to 10% of the drill holes from the 2002–2003 campaign. All reviewed folders were well organized, included collar survey data (but not original documents), drill reports, down-hole survey data, original geological logs and copies of original assay certificates. Most folders also included geotechnical logs, and density, magnetic susceptibility and PLT measurements. Some folders corresponding to earlier campaigns (1981, 1986) only included original logs.

December 2019 54

• Review of down-hole survey and assay data: AmecFW compared 2,245 down-hole survey data from original paper records with the digital records in the database and did not identify any errors. Spot checks on assay data corresponding to the reviewed folders did not reveal any differences between assay certificates and the digital database records.

• Interpretation of geology and mineralization: AmecFW reviewed the geometry of the interpreted polygons in all geological and mineralization vertical cross-sections, in order to, assess the spatial continuity and correlation to individual drill holes. The lithology cross-sections were represented at the 1:7,500 scale, and the mineralization sections were represented at the 1:2,000 scale. The sections were spaced at 50 m to 100 m. AmecFW also reviewed horizontal plans with 50 m spacing. Drill holes were represented with the projected trace, and included lithology polygons in the lithology cross-sections, and ore-zone polygons and copper and gold grades in the ore cross-sections. During the review, AmecFW did not find significant discrepancies.

• AmecFW recognized that the interpretation generally respects the data recorded in the logs, plans and cross-sections, as well as the interpretation from adjoining plans and sections, and is consistent with the known characteristics of this deposit type. The lithological and mineralization models were diligently constructed in conformance to industry standard practices.

• Core review: AmecFW reviewed selected core sections of three drill holes (FD-280, FD-296 and FD-360), and observed that the core was properly cut. The observed contacts between major units approximately matched the logged depths, although it was apparent that the boundaries between units was usually established mainly on the basis of a visual appreciation of the proportion of magnetite and garnets. Recovery in the reviewed core was excellent.

12.3. Comments on Section 12 The QPs have fully reviewed the drilling data and interpretations of geology and mineralization as well as the work done by independent third-party consultants. In the opinion of the QPs, data verification has been extensively conducted since 1988 by Vale employees and numerous consultants. No material issues have been identified in reviewing those programs. In addition, the QPs reviewed and agree with Vale’s ongoing steps and procedures to verify the quality of the data. The QPs conclude that the drilling logging and sampling procedures are appropriate for the Salobo mineralization styles, that the assay data are reliable, and that the database is reasonably free of errors. Therefore, the data is suitable to support Mineral Resource and Mineral Reserve estimation and can be used for mine planning purposes.

December 2019 55

13. Mineral Processing and Metallurgical Testing

13.1. Metallurgical Testwork Five distinct phases of testwork have been completed:

• CVRD from 1978–1981

• CVRD and Anglo American from 1986–1987

• SMSA from 1993–1998, including a pilot-plant campaign carried out at the CVRD Research Centre (CRC)

• Variability program including flotation (rougher and locked cycle tests) and comminution studies from 2003–2004; and

• Trade-off study using high-pressure grinding rolls (HPGR) for tertiary crushing as an alternative to conventional semi-autogenous grinding (SAG), from 2005–2006.

The following sections summarize the most relevant programs indicated above, as they provided the basis of the plant design criteria and/or its metallurgical performance projections.

13.1.1. Variability Tests For the 2004 variability test programs, rougher flotation tests were conducted on 251 samples obtained from drill core representing a wide range of ore grades and lithologies. Approximately 47% were classified as XMT and 41% BDX, which are the two most important lithologies. In addition, 59 locked cycle tests were carried out at Minas Gerais Technological Centre (Cetec) in Belo Horizonte, involving 30 BDX samples and 16 XMT samples. Two major metallurgical improvements were incorporated into these tests. Firstly, a reagent scheme was adopted using a blend of two collectors, A350 (potassium amyl xanthate) and A3477 (sodium di-isobutyl dithiophosphate). This resulted in improved metallurgy and stable flotation conditions. During the 1994 pilot plant trials only A350 was used, resulting in unstable flotation conditions and the evaluation of a two-stage grinding circuit. The addition of sodium sulphide during rougher/scavenger flotation was demonstrated as being important for the effective flotation of bornite, which tends to oxidize and tarnish quickly and requires higher collector addition. U.I. Minerals (Uimin) consolidated the 1994 plant trial results and the variability study results in December 2003. During this process, a number of filters were applied to the results, leaving 71 tests carried out on what was then deemed to be representative samples of the lithologies. From this exercise, an average metallurgical recovery for copper is 90.7% and a mass recovery of 18.2% was projected. A total of 177 samples were analyzed with grades

December 2019 56

above 0.4% Cu. Results showed 87.6% of the samples with a copper recovery above 90%, 10.7% of samples had a recovery between 85 and 90%, and only 1.7% of samples were anomalous with recoveries less than 85%. Copper recoveries for the 251 variability rougher tests are summarized in Figure 16. There was a direct correlation between the copper recovery and mass recovery for each lithology. XMT samples had higher mass recoveries, which could be due to either higher grades and/or the presence of magnetite in the concentrates. The average gold recovery for the deposit was 67.4% with a standard deviation of 14.4%. Approximately 64% of samples with initial gold grades above 0.4 g/t had gold recoveries of up to 70%. Gold recoveries for all the 251 variability rougher tests are summarized in Figure 17. Based on the consolidated results, equations for predicting copper and gold recoveries were developed by UIMIN, for use in mine planning and production forecasts.

• Rec Cu (%) = -0.023 / [%Cu in feed] + 0.9023 {Equation 1}

• Rec Au (%) = 0.0256 * [g/t Au] + 0.6485 {Equation 2} Uimin commented that “it should be stated that the curve fit is rather poor” for Equation 1, as exhibited by the spread of data in Figure 16. A further comment that Equation 1 is applicable mostly in the 0.6–1.5% Cu range, as head grade, was made given the test data fell outside this range. Equation 1 is derived from using the data points represented by the yellow crosses in Figure 18 and subtracting 6.5% from these values. This deduction represents the losses expected in the cleaner-scavenger tails relative to what is recovered in the rougher-scavenger. The resulting overall recovery trend appears as the orange crosses in Figure 18. The gold recovery projection represented by Equation 2 is illustrated by the red line in Figure 19 as the “best fit” for the locked-cycle tests as symbolized by the red circles. These are the equations were used in the Project justification studies and the Geology and Mine Operations departments have used them for projecting recoveries, across all lithologies, even though the pre-production testwork data had already shown differential responses for the XMT, BDX and DGRX.

December 2019 57

Figure 16 - Copper Recoveries in 2003–2004 Variability Testwork

Note: Figure from Uimin, 2004. X-axis label is the initial Cu head grade in %; Y-axis is the rougher Cu recovery in percent, SE is the southeast sector, NW is the northwest sector, and the red line is the estimated recovery curve.

Figure 17 - Gold Recoveries from Variability and LCT Testwork Programs

Note: Figure from Uimin, 2004. X-axis label is the initial Au head grade in g/t (ppm); Y-axis is the Au recovery in percent, SE is the southeast sector, NW is the northwest sector.

December 2019 58

Figure 18 - Derivation of Copper Recovery Projection, 2003–2004 Variability Testwork

Note: Figure courtesy Vale, 2014

Figure 19 - Derivation of Gold Recovery Projection from Variability Testwork


Cu-Head Grade vs Cu Recovery of ALL Data (<0.6% -1.5% Cu & Cu/S=1.75-3.30)

Variability & PP94 & LCT (Rg+Sc)Conc & PP94 & LCT Final Conc

FIGURE 2

75%76%77%78%79%80%81%82%83%84%85%86%87%88%89%90%91%92%93%94%95%96%97%98%99%

100%

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 1.55 1.60

Head Grade, %Cu

Cu

Rec

ov

ery

, %

Calc-RGHR Rec=-0.023*(1/T)+0.9673

OLD-CLNR-Rec=0.8622+0.063*LN(T)

EST-CLNR-Rec=Rghr%-6.5%

94PP-Final Conc

94PP-Rg+Sc Conc

LCT-(Rg+Sc) Conc

LCT-Final Conc

03MPP-Rg+Sc

03MPP-Final Conc

Rg+Sc & CF Au-Rec vs Feed Au-AssayALL Var.Data (0.6-1.5%Cu & Cu/S 1.75-3.25) & LCT Tests & PP94-SML04-Bilmat

FIGURE 24.D

y = 0.0408x + 0.7351

R2 = 0.0261

y = 0.0471x + 0.7116

R2 = 0.0799

y = 0.0256x + 0.6485

R2 = 0.0103

y = 0.3618x + 0.4548

R2 = 0.2805

35.0%

37.5%

40.0%

42.5%

45.0%

47.5%

50.0%

52.5%

55.0%

57.5%

60.0%

62.5%

65.0%

67.5%

70.0%

72.5%

75.0%

77.5%

80.0%

82.5%

85.0%

87.5%

90.0%

92.5%

95.0%

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6

Feed Assay, Au-g/t

Rg

+S

c &

CF

GO

LD

Rec

%

Rg+Sc Au-Rec

LCT-Rg+Sc Bench Test

LCT - CF Au-Rec

PP94-SML04-Bilmat

Linear (Rg+Sc Au-Rec)

Linear (LCT-Rg+Sc Bench Test)

Linear (LCT - CF Au-Rec)

Linear (PP94-SML04-Bilmat)

December 2019 59

13.1.2. High Pressure Grind Roll Trade-off Study (2006) The feasibility study conducted by Fluor Daniel in 2004 incorporated a conventional primary crushing circuit, a standard SAG mill/ball mill grinding circuit and a conventional copper flotation circuit. However, several unique problems with Salobo ore led to the evaluation of an alternative to standard SAG mill grinding:

• Firstly, the high magnetite content (potentially exceeding 20% at times) presents a difficulty in the SAG mill circuit due to the need to remove and crush the critical size pebbles in a pebble crusher. The use of a magnet to remove tramp steel ahead of the crusher would invariably remove magnetite pebbles with a resulting loss of the associated copper and gold values (the higher-grade copper is associated with the magnetite schists). There would then be additional design and cost implications to re-handle and process the magnetite pebbles.

• Secondly, significant variations in hardness and density were predicted for Salobo ore and conventional SAG milling circuits are sensitive to such variations, resulting in potential significant variability in mill throughput and performance.

Because of these concerns, an extensive evaluation of an alternative comminution circuit was conducted that included primary crushing, secondary cone crushing and tertiary HPGR crushing followed by conventional ball milling. Polysius conducted two separate HPGR evaluations, in 2005 and 2006. The 2005 program tested pilot ore from the G3 adit at a top feed size of 32 mm and the 2006 program tested two samples to represent typical ore for the first five years of mining and hard ore. Top feed sizes tested were 25 mm and 32 mm. The Bond ball mill work index for the hard ore sample was 21.4 kWh/t. General observations from this testwork program were that there was a decline in specific throughput as the roll speed increased and as the feed moisture content was increased. For the first five-year sample, there was an 18% reduction in specific throughput when the feed moisture content was increased from 0.1% to 4.0%. Abrasion testing and specific wear rates on all samples indicate that Salobo ore has low abrasion characteristics. Grindability tests were conducted on samples of HPGR product at <6 mm and conventionally crushed material at <6 mm of the pilot ore sample from the 2005 program. The results indicated a very similar Bond ball mill work index for both samples (19.4 kWh/t and 19.2 kWh/t, respectively), indicating no micro-fracturing of the rock and therefore no grindability advantage was attributed to HPGR.

December 2019 60

SMCC was retained as an independent reviewer of the Polysius test program and to size both the HPGR and ball mill units. Finally, Aker Kvaerner conducted a trade-off study using the results of the Polysius testwork programs and the SMCC review in 2006 After reviewing all the work, Vale decided to implement the HPGR option based on the technical and economic benefits compared to conventional SAG.

13.1.3. Mixed Ore Zone Copper Recovery Testwork A copper recovery study for the mixed ore stockpiled at the mine was commissioned in 2014. Figure 20 is indicative of the improved metallurgical response brought by much increased addition rates of the collectors currently used in the rougher flotation, as accompanied with the inclusion of sodium silicate, used as a dispersant (e.g. viscosity modifier). Another study involved the transition ore, at the boundary between the oxide cap and the sulphide ore. The objective was to find the proportion of oxidized mineralization that can be added to the fresh, sulphide-bearing bedrock without impacting the overall copper recovery in the plant. As presented in Figure 21, such work provided an indication that a mixed ore component of up to 30% could be tolerated with limited impact on the results expected with fresh material only. Work mostly carried out by the Vale Sheridan Park Research Centre (SPRC) in recent years resulted in development of a projected recovery curve for copper. SPRC also was involved in providing support towards the eventual processing of the mixed ore stockpile material. Figure 22 shows the recovery projection model, built by SPRC, and the results of testwork realized with fresh (ROM in the figure legend) and stockpiled ore samples. The equation underlying the recovery projection model is expressed as follows:

• Rec Cu(@38%Cu) = 88.5 * (1 – exp(-3.5 * [Cu in feed]) {Equation 3} Where [Cu in feed] is the copper feed grade and the resulting projected recovery is based on a standardized concentrate grade target of 38% Cu. The testwork program completed included modified reagent schemes, relative to the plant operations (changing the xanthate used from potassium amyl xanthate (PAX) to sodium isopropyl xanthate (SIPX), removing the sodium sulphide as modifier), as well as testing the addition of a desliming stage, with a cyclone, of the mixed stockpile material in an attempt to remove the most oxidized component and reduce reagent consumptions.

December 2019 61

Figure 20 - Testwork with Mixed Ore – Effect of Dispersant and Collector Dosage

Note: Figure courtesy Vale, 2015. X-axis label is addition rate of dispersant (dispersante; as sodium silicate, Na2SiO3),

in g/t (ppm); Y-axis is the addition rate of the sulphide collectors (coletor; as a potassium amyl xanthate and a sodium

dithiophosphate, in undisclosed proportions), in g/t (ppm). Legend is showing colour legend for achieved Cu rougher

recovery ranges in sensitivity area plot.

Figure 21 - Expected Metallurgy of Various Sulphide-Mixed Ore Blends

Note: Figure courtesy Vale. X-axis label indicates proportion of fresh sulphide (sulf) ore involved in the test feed, along

with proportions of mixed (mix) ore. Y-axis is Rougher Cu recovery (blue bar graphs) and estimated final copper recovery

(green bar graphs)

As illustrated in Figure 22, the ROM sample response under the modified (e.g. without sodium sulphide) or SIPX processing scheme did not present a marked advantage from the design (e.g. current plant operations) scenario. The figure is also indicative of the large degradation in the metallurgical results that would result from processing the stockpile material with the design reagent scheme and then the marginal improvement

December 2019 62

expected under the approach where the flotation feed would have been first deslimed by cycloning. Based on the current testwork knowledge, the approach retained would be to not provide the plant with more than 30% of its total feed tonnage from the mixed material accumulated on the stockpile.

Figure 22 - Test Results with Fresh and Mixed Ores with Modified Reagent Scheme


13.2. Recovery Estimates Recovery projections for copper and gold (see Section 13.5.3) are based on Equation 1 and 2, respectively and more recently, these equations have been modified on the basis of operating plant data augmented by considerable laboratory test work which has been included in the ongoing geometallurgical program. These are underlying a fixed target copper grade in concentrate of 38% Cu. These equations are currently used to project these metals’ recoveries in the Mineral Reserve estimate, cut-off grade calculations, and the life-of-mine financial model. The models are typically updated annually to take into account the most recent plant operating performance and relate that operational performance to samples tested as part of the geometallurgical program. Silver recovery was not tracked as diligently as copper and gold during the testwork phases. Silver contributes approximately 1% to the overall revenue of the project and on that basis is not reported in this Technical Report.

December 2019 63

Pre-production testwork, especially the large variability testwork program of 2003–2004, provided indications that the copper metallurgical response was variable, not only per the feed grade but also per the lithology of the ore. The adoption of a single equation to predict recovery, instead of drafting discrete equations for each lithology and using these to match the expected relative proportions provided in the mine plan as plant feed, is a simplification that may create daily discrepancies between expectations and actual results. However, this approach is likely, over monthly periods for instance, to provide a valid approximation of potential results. The shortfalls suffered during the commissioning periods of each plant are discounted. These would be the absence of sodium hydrosulphide (NaHS) as a modifier during the first six months of the Salobo I plant operations, the lack of a lime preparation system and associated safety concern related to the addition of NaHS in a low pH environment, and the high frequency of individual line shutdowns still encountered in the plant. Figure 23 and Figure 24 in Section 13.5 present the actual monthly plant results against the projections from 2013 to 2019.

13.3. Metallurgical Variability

As discussed in Section 13.2, some variability in the metallurgical results can be expected as the mixture of lithologies found in the plant feed change. Over monthly periods, the resulting blend is more likely to approach the Mineral Reserves profile and thus mitigate the variability that may be detected on a daily basis, versus projections. Introduction of mixed material above a proportion of 30% of plant feed has been shown to lead to a degradation of the flotation results. Proper blending of such material, albeit representing only 1% of the Mineral Reserves, will be required.

13.4. Deleterious Elements There are three deleterious elements of potential concern in the copper concentrate, namely fluorine, chlorine and uranium. Of these, fluorine is the most significant. In general, smelters will tend to reject concentrates with high fluorine content due to problems that result in the smelter’s sulphuric acid plants. Testwork was conducted as part of the 1994 pilot plant campaign and continued by CVRD in 1995 to evaluate the potential for acid leaching of the concentrate to reduce fluorine levels. This was apparently unsuccessful due to insufficient removal of fluorine, high dissolution of copper and difficulty in filtering the leach residue. Mineralogical examination of the ore lithologies and concentrate samples have indicated a tendency for this element to be concentrated in fluorite and silicates, mainly biotite. These gangue minerals are partially reporting to the concentrate stream mostly through partial liberation from sulphide-bearing mineral grains and mechanical entrainment in the froth phase of the flotation process. Regrinding of the scavenger concentrate, for incremental liberation, as well as cleaning in the flotation circuit with flotation columns,

December 2019 64

instead of conventional mechanically-agitated cells, was adopted to enhance liberation and provide a means for more effective froth washing, with water, of the final concentrate stream. These two approaches demonstrated a capability to improve rejection of the deleterious elements at levels acceptable to some smelters, albeit while still attracting penalties on fluorine and chlorine, for their high residual contents. Vale has secured contracts with smelters able to accept the copper concentrate, with an average fluorine content of about 2,000 ppm, and a maximum content of 4,000 ppm. Penalties are charged though starting below the actual content. These smelters also placed the maximum acceptable chlorine content at 1,200 ppm, but with a penalty drawn at the 550–650 ppm. Uranium, is present in the copper concentrate. Recent annual averages of uranium levels are between 50 to 60 ppm. The specification limit is 60 ppm. There is a strong correlation between uranium in the concentrate to uranium in the feed. Operational procedures are being implemented to accurately forecast the uranium grade in the long term and short term models to enable blending of plant feed from the mine. Since concentrate lots are segregated by grade (lower, medium and high grades) at the Parauapebas transfer shed and at the port of São Luis, blending of out-of-specification concentrate is possible, should it ever be necessary.

13.5. Actual Plant Results versus Budget and Model Projections Considering the accumulated historical data from the plant operations, it is relevant to assess how current results match expectations from the design stage and how they are integrated in the prediction of future results, both from the aspects of the metallurgy and of the throughput capability of the plant.

13.5.1. Historical Metallurgical Results Figure 23 and Figure 24 are indicative of the monthly plant performance from January 2013 to December 2019 in relation to the predicted recoveries based on testwork data and Equation 1 and Equation 2. Section 13.2 offered an overview of some of the events that may have contributed to the differentials shown in these figures over specific periods, with copper having apparently suffered more shortfalls than gold as a result of these. Geometallurgical programs were initiated in 2014 and continued in 2018-2019 as described in Section 13.5.3. The testing and data analysis of historical plant operating performance produced revised equations for copper and gold recovery which were used to predict copper and gold production based on budget values of tonnes processed and copper and gold grades delivered to the process plant in the budgeting for 2019 production. The 2019 recovery equations are:

December 2019 65

Cu recovery % = -5.5884 * (1/Cu in feed %) + 92.8945 Au recovery % = (43.661 * Au in feed g/t) + 44.354 Figure 25 to Figure 28 present in more detail the recent metallurgical results in terms of the 2019 monthly actual data. The 2019 copper recovery target was 86.95% in the budget, and the actual copper recovery in 2019 was 86.91% (Figure 23). Using Equation 1, the actual copper grade in 2019 of 0.969% would have predicted a copper recovery of 88.0% so the actual copper recovery was 98.8% of the predicted value. A comparison to the equations used to budget the 2019 production gives a predicted copper recovery of 87.13% with the actual 2019 copper grade of 0.969%. This validates the improvement in the accuracy of the copper recovery model for 2019 in that the actual recovery was 99.75% of the predictive model recovery. A very close result. The 2015 - 2019 results for copper recovery represent a marked improvement both from 2014, which may have been negatively influenced by the commissioning of Salobo II, and from 2013, at 81.4% and 80.9%, respectively (Figure 25). The 2019 gold recovery target is was 70.5% in the budget, in comparison to 74.44% achieved (refer to Figure 23). Using Equation 2, the 0.684 g/t Au head grade budgeted for the full year would have called for a projected recovery of 66.6%. The actual gold recovery of 74.44% is 112% of the predicted value using Equation 2. If gold recovery predicted by the 2019 model and the actual 2019 gold grade of 0.684 g/t is calculated, it results in a gold recovery of 74.23%. This again validates the improvement in the accuracy of the gold recovery model used in 2019 in that the actual gold recovery was 100.28% of the predictive model. Gold recovery from 2013 - 2019 has been consistent at 65 - 75%, generally above budget (Figure 24 and Figure 26) with the lower recoveries the result of frequent process plant stoppages during the production ramp-up of both Salobo I and II with higher grades in 2013-2014 offsetting these process interruptions.

December 2019 66

Figure 23 - Actual versus Projected Monthly Plant Copper Recovery

Figure 24 - Actual versus Projected Monthly Plant Gold Recovery

60.0

65.0

70.0

75.0

80.0

85.0

90.0

95.0

2013 2014 2015 2016 2017 2018 2019

%

Budget Cu recovery (%) Actual Cu recovery (%)

30.0

40.0

50.0

60.0

70.0

80.0

2013 2014 2015 2016 2017 2018 2019

%

Budget Au recovery (%) Actual Au recover (%)

December 2019 67

Figure 25 - Historical Plant Performance – Cu Recovery

Figure 26 - Historical Plant Performance – Au Recovery

80

.9

81

.4

86

.8

87

.5

88

.3

86

.1 88

.2

87

.3

86

.0

87

.6

83

.1 87

.4 89

.4

87

.0

84

.9 86

.6

88

.3

86

.0

86

.9

50.0

60.0

70.0

80.0

90.0

100.0

20

13

20

14

20

15

20

16

20

17

20

18

Jan

Feb

Mar

Ap

r

May Jun

Ju

l

Au

g

Sep

Oct

No

v

Dec

20

19

%Copper Recovery

65

.2

64

.2 67

.5

69

.3

69

.0 71

.7

73

.0

74

.4

72

.5

74

.0

70

.1 74

.7 77

.0

76

.5

72

.6 74

.7

76

.1

76

.0

74

.4

30.0

40.0

50.0

60.0

70.0

80.0

90.0

20

13

20

14

20

15

20

16

20

17

20

18

Jan

Feb

Mar

Ap

r

May Jun

Ju

l

Au

g

Sep

Oct

No

v

Dec

20

19

%

Gold Recovery

December 2019 68

Figure 27 - Historical Plant Performance – Concentrate Cu Grade

Figure 28 - Historical Plant Performance – Concentrate Au Grade

Equation 1 was derived on the premise of achieving a final copper concentrate grade of 38% Cu. This concentrate grade is used in the budget and financial model, and by the operations personnel to decide whether to pull the flotation circuit harder or not. As shown in Figure 27, a very good performance from the plant has been achieved in this regard and should be able to be upheld in the future. The gold grade in the copper concentrate has been consistently between 19 and 25 g/t with variations being attributable to different gold to copper ratios in the feed.

39

.4

38

.5

38

.6

39

.5

38

.8

37

.8

38

.2

37

.3

37

.7

37

.8

38

.1

37

.6

36

.3

36

.0

36

.8

37

.2

37

.4

36

.4

37

.2

0.0

10.0

20.0

30.0

40.0

50.0

20

13

20

14

20

15

20

16

20

17

20

18

Jan

Feb

Mar

Ap

r

May Jun

Ju

l

Au

g

Sep

Oct

No

v

Dec

20

19

%Concentrate Copper Grade

21

.9

19

.5

19

.4 22

.2

21

.6

22

.1

21

.1

26

.6

23

.2 25

.6

22

.3

20

.8

21

.4

22

.2

21

.6

21

.2

21

.5 23

.6

22

.5

0.0

10.0

20.0

30.0

20

13

20

14

20

15

20

16

20

17

20

18

Jan

Feb

Mar

Ap

r

May Jun

Ju

l

Au

g

Sep

Oct

No

v

Dec

20

19

g/d

mt

Concentrate Gold Grade

December 2019 69

13.5.2. Historical Plant Availability and Utilization The throughput capability of the plant has been increasing over time through improvements in both the plant availability and utilization. These have both been improving over the initial few years of operation when ramp-up and commissioning of the Salobo I and Salobo II circuits was occurring. The improvements are noted in Figure 29 and Figure 30.

Figure 29 - Salobo Plant Historical Availability

Figure 30 - Salobo Plant Historical Operational Utilization

67

.2 75

.1 81

.7

83

.5 89

.1

89

.8

89

.6

77

.7

92

.1

93

.4

77

.5 85

.6 93

.6

91

.9

78

.1

92

.5

92

.5

89

.1

87

.9

0.0

20.0

40.0

60.0

80.0

100.0

20

13

20

14

20

15

20

16

20

17

20

18

Jan

Feb

Mar

Ap

r

May Jun

Ju

l

Au

g

Sep

Oct

No

v

Dec

20

19

%

Plant availability (%)

91

,2

90

,7

94

,4

98

,1

98

,4

99

,0

98

,7

96

,4

97

,2

94

,2

98

,0

98

,7

98

,5

99

,1

96

,8

96

,6

97

,3

96

,1

97

,3

0,0

20,0

40,0

60,0

80,0

100,0

20

13

20

14

20

15

20

16

20

17

20

18

Jan

Feb

Mar

Ap

r

May Jun

Ju

l

Au

g

Sep

Oct

No

v

Dec

20

19

%

Plant utilization (%)

December 2019 70

The onus for achieving incremental processed tonnage is thus mostly placed on the maintenance group, as it is tasked with generating the availability improvement expected from the equipment, and on the process control group, with control tuning and strategic modifications. The 2020 production plan requires that the availability target gradually increases to 88.9% by 2020. With the 2017-2019 averaged achieved availability of 88.9%, it has been demonstrated that the plant can run at the required availability.

13.5.3. Recovery Projections The estimates used for 2020-2023 plant recoveries of copper and gold have been based on recent plant operating performance and the most recent geometallurgical testing programs and are shown in Table 11.

Table 11 - Processing Recovery Assumptions (2020-2023)

Metal 2020 Total Recovery to All Concentrates (%)

Copper -3.154*(1/A) + 90.6108*(-0.0745*B + 1.2321) - (3.96* C)

Gold -59.421*(D^3) + 152.37*(D^2)-103.9*D + 93.388*(-0.0745*B + 1.2321)- (2.21* C)

Metal 2021-2023 Total Recovery to All Concentrates (%)

Copper -3.154*(1/A) + 90.6108*(-0.0745*B + 1.2321) - (3.96* C)

Gold -59.421*(D^3) + 152.37*(D^2)-103.9*D + 91.815*(-0.0745*B + 1.2321)- (2.21* C)

Notes: Factors A=Feed Cu grade (%), B=Wet Plant feed rate (kmt/hour), on the basis of two lines of operation C=Percentage in mass of transition ore (%) and D=Feed Au grade(g/t)

The gold recovery formula for 2020 gives slightly higher recoveries as compared to the formula used in 2021-2020. This is based on the improved gold performance seen in 2019 and the geometallurgical testing in 2019 on samples forecast to be in the 2020 mine plan. It is planned to test 2021 samples in 2020 so the gold recovery model could be updated as plant operational data and test work continue to be evaluated. Ongoing improvements to plant operation are being continuously investigated with some of the recent initiatives being described in Section 13.6. In 2021 only, a very small amount of transitional ore is planned to be processed (<0.0025% of the total feed tonnage), with this low amount of this ore type, the copper and gold recoveries are not materially affected. Estimates of plant recoveries for 2024-remaining LoMP are shown in Table 12 below.

Table 12 - Processing Recovery Assumptions (2024-LoMP)

Metal Total Recovery to All Concentrates (%)

Copper (-2.5362*1/%Cu in feed ) + 90.674

Gold (1.0173*RecCu)-20.357

Equations to project copper and gold recoveries used a large dataset from testwork. Within the 0.6–1.5% Cu range of the data points retained for a regression-analysis, the resulting equations are fairly insensitive to the actual feed grade encountered.

December 2019 71

The “noisy” set of testwork data may have been divided by lithologies for analysis, in order to use this additional information as a projection tool. More recent testwork confirmed a better copper response for XMT than for BDX, for example. The potential variability of gold, as it relates to lithology, is less clear. The recovery projection equations may therefore be sufficient for predicting results over longer-term periods (e.g. yearly, maybe monthly), but may not be adequate for applying a daily target to the plant operations since variations in the lithological make-up of the plant feed over such a short period may have called for different target recoveries than indicated by the equations. A geometallurgical program initiated in June 2014 is expected to allow for building a database from which refinements to the projection tools, if deemed valid, could be made. Results achieved to date from operations indicate that metallurgical targets are reasonable and fairly aligned with the projection tools in place, once consideration is made for ramp-up effects. The geometallurgical routine process was improved in 2019 based on the following actions:

• Improvement in the in-house system for monitoring the plant feed and geometallurgical performance in real time.

• Implementation of tests to optimize flotation regulation. In 2019, the long-term geometallurgical program was focused in studying the geometallurgical behavior of the ore inside the 2020-2024 mining plan, to complement the study that was done with the ore samples from the 2019 mining plan. These samples were gathered from the long-term drilling. Rougher, open cleaner and LCT flotation assays were performed with 12 global samples from 2019 plan and 20 global samples from 2020-2024 mining plan. The 2019 global samples generated 24 subsamples, while 2020-2024 global samples generated 41 subsamples for rougher and cleaner assays. The LCT tests showed that the samples had higher copper and gold metallurgical recoveries than those defined in the Salobo project (87%). Copper and gold recoveries were respectively 90.8% and 67.0% for 2019 samples and 89.2% and 64.9% for the 2020-2024 samples. The concentrates obtained in the LCT assays with the 2020-2024 samples had lower payable metal content (Cu, Au and Ag) and higher deleterious contents (F, Cl and U) compared to the chemical quality of the products obtained with the 2019 samples. The average F, Cl and U grade in the concentrate obtained from the 2020-2024 samples were respectively 1580 ppm, 1008 ppm and 56 ppm, while the concentrate obtained from the 2019 samples were 1160 ppm (F), <615 ppm (Cl) and 46 ppm (U).

December 2019 72

The throughput (t/h) and specific energy consumption values in HPGR and ball milling (kWh/t) ranged from 2,890 to 3,500 t/h (median: 3,125 t/h), from 1.59 to 2.05 kWh/t (median: 1.87 kWh/t) and from 16.67 to 20.30 kWh/t (median: 18.74 kWh/t), respectively. The values obtained in the simulations showed variations consistent with those practiced in the industrial plant, as shown by data from the operation in 2018. The main goal of the long-term geometallurgical study is to model recoveries, concentrate deleterious and energy consumption in the long-term block model, which is the objective for 2020.

13.6. Process Plant Optimization Figure 31 and Figure 32 show the as constructed Salobo Mineral Processing Facility and process flowsheet.

Figure 31 - Salobo Line I and II Mineral Processing Facility

December 2019 73

Figure 32 - Salobo Flow Sheet

Changes and accomplishments at the Salobo processing plant in 2018 and 2019 include the following:

Salobo I

Salobo II

December 2019 74

• Collectors and frother flotation reagents were changed to improve the copper recovery.

• Improvements in the tramp metal extraction system (guard magnets), increasing the HPGR’s rolls lifespan.

• Power increase on vertical mills to reduce uranium flotation.

• Aeration system improvement in the flotation columns to increase recoveries.

• Improvement in the feed system of the secondary crusher, increasing the productivity and feed rate.

• Replacement of the mill cyclones, reducing the circulating load and increasing the milling rate.

• Improvement in the power system of the thickeners, allowing handling of the high-grade ores without losses due to overload.

• Implementation of OCS (Optimizer Control System) for the flotation.

• Implementation of DFR (Direct Flotation Reactor) in the cleaner of Salobo

concentrate line 2.

• Revitalization of mechanical components of flotation cells.

• Replacement of pipes for larger diameter ones in the concentrate lines.

• Repowering of pump system in the mill line 2.

The geometallurgical routine process was improved in 2018 and 2019 based on the

following actions:

• Metal recovery reconciliation process which compares the estimated recovery by the bench tests with the actual plant recoveries.

• Implementation of an in-house system for monitoring the plant feed and geometallurgical performance in real time.

• Expansion of the Salobo Operations geometallurgical laboratory and acquisition of the new equipment, improving the turnaround time.

• Start of the long-range geometallurgical model through the metallurgical analysis and characterization of samples from the current long-range drilling campaign.

• Improvement in the in-house system for monitoring the plant feed and

geometallurgical performance in real time.

• Implementation of tests to optimize flotation regulation.

December 2019 75

13.7. Comments on Section 13 In the QPs opinion, the testwork is appropriate to be the basis for the Salobo concentrator design. This is evidenced by the recent metallurgical performance of the commercial plant being very close to that predicted by models derived from the results of the extensive testwork and ongoing modifications to the predictive models for copper and gold recovery. Throughput and recovery ramp-up assumptions for Salobo III have been incorporated into the production profiles for 2022 and 2023 and are based on the achievements seen in the Salobo II ramp-up.

14. Mineral Resource Estimates Mineral Resource estimation is completed by João Dirk under the supervision of Marcos Dias Alvim, both Vale employees. The estimates are prepared according to the 2014 CIM Definitions Standards and the 2003 CIM Best Practice Guidelines. Mineral Resources that are not Mineral Reserves do not have demonstrated economic viability. There has been insufficient exploration to classify the Inferred Mineral Resources as an Indicated or Measured Mineral Resource. The extent to which further exploration may result in upgrading them to an Indicated or Measured Mineral Resource category is uncertain at this time. An infill drilling program commenced in 2017 and continues to date (Table 13), targeting areas of lower density drilling with the intent of upgrading Inferred Mineral Resources. The infill drill program target is to drill 10,000 m per year to improve the long-range geological model contours, grade estimation, uncertainty, predictability and classification for the 5-year plan. Approximately 35,500 m was drilled until September 2019 (Table 13). Blasthole information is also used in updating the long-range geological model contours in mined-out areas.

Table 13 - Summary of Salobo Infill Diamond Drilling Campaign

Year Exploration (m) Definition (m) Total (m)

2017 1,565.9 13,265.0 14,830.9

2018 4,300.2 12,674.5 16,974.7

2019 1,370.0 10,159.0 11,529.0

Total 7,236.1 36,098.5 43,334.6

At the end of 2018, 25,000 m had been drilled but only 5,000 m had been analyzed due to sample preparation and analysis issues at the time. Since Salobo did not have a drill program since 2003, it took some time to ramp up the processes required to manage the core. The core shed was cleaned and prepared to receive the equipment. Saw machines, weighing scales and other equipment were purchased and the Salobo

December 2019 76

Operations laboratory had to build a separate preparation line to prepare the new core. The 5,000 samples were included in the 2018 updates are located mainly in the phase 4 of the mining plan. In September 2019, all the backlog from 2017 and 2018 was logged, sampled, and analyzed. At the end of 2019 there was a sampling backlog of 5,000 m from the 2019 drilling. The December 31, 2019 block model update included 22,000 new samples. A deep drilling campaign began in 2017 and at the end of 2019 five holes had been completed and a sixth was in progress. Of the five holes completed one still had assay results pending. All holes encountered mineralization below the current resource pit. A deep exploration study will be completed in 2020 with the goal of defining the next deep exploration targets.

14.1. Introduction Mineral Resource modeling for Salobo utilizes drilling data, enhanced knowledge of metallurgical processing, geology and mineralization, and refined interpolation parameters. The geologic and Mineral Resource models were constructed using GEMS™ and Isatis® software. The estimated Mineral Resources are then converted to Mineral Reserves using long term mine planning techniques and quoted above a cut-off grade of 0.253% Cu equivalent (CuEq). Only diamond drill hole composites form the database and are considered in building the Mineral Resource model for the Salobo deposit. Mineral Resources were classified as Measured, Indicated and Inferred in accordance with 2014 CIM Definition Standards. No Inferred Mineral Resources are converted to Mineral Reserves. Table 14 summarizes the estimated Mineral Resources at the Salobo Operations as of 31 December, 2019.

Table 14 - December 31, 2019 Mineral Resource Estimates

Grades Contained Metal Classification M Tonnes Cu % Au g/t Cu (M lb) Au (M oz)

Measured 1.2 0.67 0.42 17 0.02 Indicated 192.3 0.61 0.31 2,586 1.9

M&I 193.5 0.61 0.31 2,603 1.9

Inferred 176.1 0.59 0.29 2,281 1.6

Notes: 1. Mineral Resource estimates were prepared by Mr Joao Dirk V. Reuwsaat, a Vale employee. The Qualified Person for

the Mineral Resource estimates is Mr. Marcos Dias Alvim, P.Geo., FAusIMM(CP), Long Term Planning Manager, South Atlantic Operations, Vale Base Metals.

2. Mineral Resources are classified as Measured, Indicated and Inferred Mineral Resources and are based on the 2014 CIM Definition Standards.

3. Mineral Resources are exclusive of Mineral Reserves. 4. Mineral Resources that are not Mineral Reserves do not have demonstrated economic viability. 5. Mineral Resources are reported above a copper equivalent cut-off of 0.253%, assuming $1,290 per ounce gold and

$3.18 per pound copper. 6. Tonnages are rounded to the nearest 100,000 tonnes; grades are rounded to two decimal places. 7. Contained copper is reported as Imperial pound units and contained gold as troy ounces.

December 2019 77

14.2. Geological Interpretation Mineral resource estimates are based on a three-dimensional computer block model utilizing GEMS™ software. Horizontal and vertical block sizes were chosen to adequately model the geometry of mineralized zones, and to approximate the selective mining unit (SMU) based on the proposed mining fleet. The resource block model was estimated using a regular block model with block sizes of 30 m by 30 m by 15 m. The geologic models for lithology and mineralization were produced by Vale geologists and were based on their experience of the geological features of the deposit, including structure, hydrothermal alteration minerals, lithologies and mineralization. The following two zones were interpreted:

• Low grade grading between 0.2 to 0.6% Cu, corresponding to the structurally controlled alteration halo consisting primarily of the BDX unit; and

• High grade above 0.6% Cu, consisting primarily of the XMT unit. In general, a minimum drill-intercept width of 8.0 m was used to define the two mineralized zones and internal barren or weakly mineralized zones. Isolated intervals below 0.2% Cu are included in mineralized zones to provide continuity of geometry from hole to hole and section to section. Narrow isolated intervals grading above 0.2% Cu are generally not interpreted as mineralized zones. Other than the intersection of the diabase dyke, little or minimal structural disturbance is observed within the Salobo deposit and for the purpose of mineral envelope modelling has not been considered. The presence of any faulting is noted but due to the orientation of these faults the operation has taken the decision not to model these at this time as they are not thought to materially impact the estimation model. Mineral Resource estimation was undertaken by applying knowledge of the deposit and understanding of local variations within each domain that control the spatial grade variation. This is further investigated by testing the search parameters to arrive at the most robust estimate. The Ordinary Kriging (OK) method was used to estimate block grades. Estimation for all variables was performed using a three pass OK approach by estimation domain. One additional pass was performed for domains to allow for the estimation of all blocks. A block discretization of 5 m by 5 m by 5 m was adopted for the 15 m block bench height. Salobo was divided in two sectors from west to east to account for changes in the orientation and style of the mineralized zone along strike (Figure 33). Polygonal shapes were used to create solids to code the other sectors, based on level plan views of the mineralized zones.

December 2019 78

Figure 33 - Salobo Deposit Sectors


Triangulated solid models were also created for each of the waste rock types using generalized geologic sections and level plans. In the case of late-stage, unmineralized units such as mafic dykes, the solids are used to overwrite mineralization codes in the block model. Figure 34 shows the grade shells used in the block modeling in vertical section view.

December 2019 79

Figure 34 - Cu Grade Shell Model – Section 1100SE (looking west)

14.3. Domaining The estimation domains are based on units defined for total copper and gold and are the result of a combination of sector, ore code and weathering variables. The subdivision between NW and SE sectors is defined by difference in deformation and hydrothermal alteration. The existence of a diabase dyke with strike N70°E defines the border between sectors. Another relevant aspect is the dip of the lithological units. In the SE sector, the dip is subvertical to southwest and in the NW sector the dip is subvertical to northeast. Block model domain and zone codes as they relate to sectors and grade groups are shown in Table 15 and Table 16.

December 2019 80

Table 15 – Block Model Domains Codes

Oxide Sulphide

Sector Description Code Description Domain

Southeast

Low Grade -Saprolite 1101 Low Grade -Fresh Rock 1103

Low Grade -Semi-Weathered 1102

High Grade -Saprolite 1201 High Grade -Fresh Rock 1203

High Grade -Semi-Weathered 1202

Northwest

Low Grade -Saprolite 2101 Low Grade -Fresh Rock 2103

Low Grade -Semi-Weathered 2102

High Grade -Saprolite 2201 High Grade -Fresh Rock 2203

High Grade -Semi-Weathered 2202

Table 16 - Block Model Zone Codes

Code Description

1 Oxidized (SAP) 2 Transition Ore (ZTR) 3 Sulphide Ore (RFR)

Figure 35 and Figure 36 plot mean grades versus standard deviation for copper and gold respectively (Micon, 2013). The plots show the High Grade Domains (1203 and 2203) and Low Grade Domains (1103 and 2103) having very distinct populations for both copper and gold which supports the domaining strategy.

Figure 35 - Domain Definition for Copper

December 2019 81

Figure 36 - Domain Definition for Gold

14.4. Statistical Analysis

14.4.1. Raw Assay Statistics Table 17 and Table 18 show the raw assay statistics by lithological units for copper and gold respectively. The coefficient of variation (CV) is a ratio of the standard deviation to the mean, and is useful measurement of relative grade variability. Higher CV values indicate higher variability. The BDX, DGRX and XMT are the most important lithologies in the deposit economics. CVs for copper for these three lithologies are all quite low being below 1.5. Gold CVs for the BDX and DGRX lithologies are relatively high (>3.0) but the XMT is just over 1.5.

December 2019 82

Table 17 - Copper Assay Statistics by Lithology

Litho Description # Samples Min Max Mean Median St Dev CV

All All 93,532 0.00 30.00 0.59 0.32 0.84 1.42

BDX Biotite-Granade-Schist 45,989 0.00 20.00 0.55 0.35 0.67 1.22

BIF Banded Iron Formation 239 0.01 3.00 0.34 0.18 0.43 1.27

COB Cover 122 0.06 1.40 0.40 0.40 0.17 0.44

CX Clorite-Schist 580 0.00 2.45 0.36 0.20 0.41 1.13

DB Diabase 1,212 0.01 3.20 0.13 0.02 0.28 2.18

DGRX Garnet-Grunerite Schist 11,332 0.00 30.00 0.55 0.30 0.81 1.49

GM Granitoid 5,624 0.00 8.90 0.17 0.06 0.38 2.20

GR Granite 362 0.01 3.29 0.22 0.07 0.42 1.92

HD Hydrothermalite 5,292 0.00 25.50 0.41 0.17 0.82 1.99

ML Mylonite 4,661 0.00 12.10 0.35 0.18 0.51 1.46

MTB Metalvolcanic Basic 1,489 0.01 14.80 0.39 0.15 0.76 1.95

QML Quartz-Mylonite 4,392 0.01 8.90 0.21 0.09 0.40 1.87

XMT Magnetite Schist 11,737 0.01 15.00 1.46 1.12 1.27 0.87

Table 18 - Gold Assay Statistics by Lithology

Litho Description # Samples Min Max Mean Median St Dev CV

All All 93,292 0.00 67.27 0.33 0.10 1.03 3.09

BDX Biotite-Granade-Schist 45,912 0.00 58.37 0.28 0.12 0.94 3.34

BIF Banded Iron Formation 239 0.01 1.11 0.09 0.04 0.14 1.68

COB Cover 122 0.01 1.43 0.10 0.07 0.14 1.38

CX Clorite-Schist 580 0.00 31.41 0.22 0.07 1.33 6.18

DB Diabase 1,212 0.01 1.46 0.03 0.01 0.11 3.69

DGRX Garnet-Grunerite Schist 11,291 0.00 31.45 0.24 0.09 0.64 2.61

GM Granitoid 5,623 0.00 26.55 0.08 0.02 0.45 5.39

GR Granite 362 0.00 1.52 0.07 0.01 0.18 2.59

HD Hydrothermalite 5,292 0.00 67.27 0.18 0.04 1.23 7.01

ML Mylonite 4,661 0.00 34.37 0.15 0.06 0.61 3.99

MTB Metalvolcanic Basic 1,489 0.01 15.64 0.16 0.03 0.69 4.29

QML Quartz-Mylonite 4,392 0.01 7.77 0.06 0.01 0.22 3.45

XMT Magnetite Schist 11,616 0.01 54.65 1.05 0.60 1.66 1.59

14.4.2. Compositing The sample intervals are generally 1.0 m and adhere to breaks such as geological contacts, faults and obvious changes in metal content. Two metre down-hole composites were created for statistical and geostatistical analysis and block grade interpolation. This composite length was chosen to provide the greatest amount of detail for estimating mineralized zones and to provide greater flexibility in dilution control and number of composite samples used for interpolation. Samples were composited in Isatis® to 2.0 m using the Regularization tool. The compositing process considered breaks in the presence of non-assayed intervals or in ore/waste contacts. If a sample interval is not assayed, it is not used in the calculations of the composite value.

December 2019 83

14.4.3. Domained Composite Statistics Subdividing a deposit into the domains segregates areas of common grade populations which normally reduces grade variability. Compositing also reduces variability by smoothing the effects of any short intervals of anomalous grades over a wider common interval. Table 19 details the composite statistics for copper and gold grouped according to the interpolation domains described above. CV values for both copper and gold are reduced from the raw assay statistics shown in Table 17 and Table 18. The level of grade variability indicated by these CVs supports the use of OK for block grade interpolation with proper attention to the handling of outliers.

14.4.4. Outlier Analysis Outlier analysis of the original assays included a statistical review of the grade populations and a visual review of the location high grades outliers. The statistical review examined CVs as discussed above and plotting of scatter and histograms. The plots are used to identify the point at which high grade tails separate from the main populations through the upper percentiles. The grades deemed as outliers were spatially reviewed to determine if they were truly outliers or pockets of anomalously high grades. In some cases, the core was examined to better understand the local context of the high grades. The resulting outlier strategy included both grade capping and outlier restriction and was done in two steps: 1) grade capping of assays during compositing (Table 19); and 2) outlier restriction on the composites during block grade estimation (Table 20).

Table 19 - Grade Capping Levels

Copper Gold

Domain Threshold

(%) Cumul Freq

Original CV

Capped CV

Threshold (ppm)

Cumul Freq

Original CV

Capped CV

1103 7.40 > 99% 1.07 1.045 15.64 > 99% 4.069 2.53

1203 10.50 > 99% 0.89 0.85 17.95 > 99% 1.71 1.44

2103 4.70 > 99% 1.11 1.01 16.12 > 99% 5.31 3.23

2203 6.90 > 99% 0.73 0.72 17.37 > 99% 1.88 1.66

Table 20 – Outlier Restriction

Passes 1 & 2 Passes 3 & 4

Copper Gold Copper Gold

Domain Threshold

(%) Distance

(m) Threshold

(ppm) Distance

(m) Threshold

(%) Distance

(m) Threshold

(ppm) Distance

(m)

1103 2.58 35 1.75 30 2.58 50 1.75 50

1203 2.40 10 3.01 20 2.40 20 5.30 30

2103 2.50 20 1.60 40 2.50 40 1.60 80

2203 4.50 20 4.84 40 4.50 40 4.84 80

14.5. Continuity Analysis Experimental grade correlograms were modelled from the composited drill hole data for copper, gold, specific gravity, silver, carbon, sulphur, molybdenum, fluorine and uranium

December 2019 84

for the Low and High Grade domains. The nugget effect was obtained using “down the hole” correlograms. The Low and High Grade domains were combined to produce larger datasets for analysis. Table 21 shows the resulting correlogram models for copper and gold.

Table 21 – Salobo Variography

Element Domain Direction Rotation ADA Nugget S1 R1 S2 R2 S3 R3

Cu 1103 & 1203

1 140 30 55 700

2 0 0.21 0.5 35 0.16 130 0.13 700

3 90 15 60 90

Cu 2103 & 2203

1 110 20 100 500

2 0 0.21 0.48 35 0.21 110 0.1 600

3 90 20 70 100

Au 1103 & 1203

1 140 30 50 550 2 0 0.42 0.27 20 0.2 80 0.11 700 3 90 10 25 100

Au 2103 & 2203

1 110 35 110 400

2 0 0.42 0.42 15 0.1 80 0.06 500

3 90 20 60 80 Note: ADA = azimuth dip azimuth, S = sill and R = range

14.6. Block Modeling

14.6.1. Dimensions A partial (percent) block model was generated in GEMS™ with the dimensions outlined in Table 22. The model is rotated 21.27º clockwise so that block model X axis lies along the general strike of N111.27E. Blocks were assigned percent volumes using the four domain wireframes (1103, 2103, 1203 and 2203).

Table 22 - Block Model Origin

Axis Origin* Block size

(m) No of

Blocks Model Extension

(m)

X 548,540 30 174 5,220 Y 9,359,800 30 75 2,250 Z 547.5 15 67 1,005

*Origin in GEMS is defined as top left corner of the block model

14.6.2. Boundary Conditions Table 23 details the boundary conditions that were applied to the model with respect to the sharing of composites between domains. Between the Low Grade domains (1103 and 2103) soft boundaries were used during the estimation of Cu, Au, Ag, S, Mo and U grades and hard boundaries for density and all other elements. Hard boundaries were used between High Grade domains 1203 and 2203 in estimating density and all grades. In recent years it has been recognized that there is a high degree of continuity vertically between the fresh rock and semi-weathered. For this reason, composites from the semi-weathered low grade domains (1102 and 2102) were included with the composites for estimation of the fresh low grade domains (1103 and 2103). For high grade domain 1203 composites from the high grade semi-weathered domain 1202 were included and for high grade domain 2203 composites from 2202 were included.

December 2019 85

Table 23 - Boundary Conditions

Domain Cu Au Ag S Mo U Density F C

1103 1103 & 2103 &1102 1103 & 1102

2103 2103 & 1103 & 2102 2103 & 2102

1203 1203 & 1202

2203 2203 & 2202

14.6.3. Block Estimation Block model interpolation of all variables was done using the Isatis® software applying the various parameters described above. Block grades were estimated using the OK algorithm and the parameters described in the sections above. Blocks were estimated during three successive passes of OK and then a final fourth pass was done to estimate blocks that were not informed during the first three passes. Blocks estimated during the fourth pass were not included in the classified resources but rather are intended for use in defining exploration targets during future drill programs. Block estimation is done on a 15 m x 15 m x 15 m block model with discretization set to 5 x 5 x 5 discretization points. The resulting block estimates are then reblocked to 30 m x 30 m x 15 m blocks and then imported into GEMS™. The nuggets, nested sills and ranges shown in Table 21 were applied in the OK interpolation, as well as, the outlier restrictions and boundary conditions described above. The rotation angles and dimensions of the search ellipses were based on the correlograms. The same angles were used for all four OK interpolation passes but increasingly larger search ellipse dimensions were used with each pass. Table 24 details the search ellipse dimensions for copper and gold.

Table 24 - Search Ellipse Dimensions – Copper & Gold

Domain Direction

(ADA)

Pass 1 (m) Pass 2 (m) Pass 3 (m)

Major Semi-Major Minor Major

Semi-Major Minor Major

Semi-Major Minor

SE Sector

140

60 60 60 110 130 90 500 700 400 0

90

NW Sector

110

60 60 60 110 130 90 500 700 400 0

90

Density values were assigned according to lithology for the blocks outside of the Low and High Grade domains. The average density for each lithology was based on the mean of the SG measurements for each specific lithology. Block density within the Low and High Grade Domains was estimated by OK of the SG measurements.

December 2019 86

14.6.4. Classification Coding Classification of blocks was initially assigned according to the pass in which the blocks were estimated. Blocks estimated during pass 1 were coded as Measured, pass 2 as Indicated and pass 3 as Inferred. Subsequently, this automated classification was adjusted to recode any anomalous blocks situated in areas of common category.

14.6.5. Model Validation The following methods were used to validate the block grade estimates:

• Global mean comparison of mean composite, OK and Nearest Neighbor (NN) block grade estimates

• Visual inspection of the composite and block model grades

• Swath plots of OK versus NN block grades on a series of sections and plans throughout the deposit

Table 25 shows a global comparison mean, standard deviation and CV between the composites, OK block grades and NN block grades. For both copper and gold the OK block grades compare very well to the composites and NN block grades. For domain 1203 the mean OK copper and gold grades are equal to the NN grades. For domain 2203 mean OK copper and gold grades are 1.6% and 3.4% below the NN grades, respectively.

Table 25 - Global Mean Analysis

Composite Data OK block grades NN block grades

Element Domain Count Mean St Dev CV Count Mean St Dev CV Count Mean St Dev CV

Cu %

1103 18,285 0.45 0.35 0.8 131,177 0.47 0.18 0.4 131,177 0.46 0.2 0.4

1203 12,162 1.21 0.72 0.6 55,825 1.22 0.28 0.2 55,825 1.22 0.46 0.4

2103 6,475 0.47 0.4 0.9 61,052 0.47 0.14 0.3 61,052 0.45 0.21 0.5

2203 4,701 1.32 0.85 0.6 30,794 1.27 0.33 0.3 30,794 1.29 0.52 0.4

Au ppm

1103 18,287 0.2 0.25 1.3 131,177 0.2 0.2 0.6 131,177 0.21 0.14 0.7

1203 12,161 0.82 0.78 1 55,825 0.74 0.74 0.5 55,825 0.74 0.54 0.7

2103 6,476 0.16 0.22 1.4 61,052 0.17 0.17 0.6 61,052 0.17 0.14 0.9

2203 4,701 0.6 0.69 1.2 30,794 0.56 0.56 0.6 30,794 0.58 0.42 0.7

Visual inspection of block and composite grades on plans and sections showed good correlation between the input data and output values. No obvious discrepancies were noted. To test the local estimation accuracy for each domain, swath plots were created comparing OK block grades versus NN block grades. These plots consist of narrow slices generated through the deposit along northing, easting and elevation directions. All domains show good correspondence between the OK and NN block estimates for both copper and gold with the OK grades being somewhat smoother as expected from the effects of the kriging interpolation. Portions of the graphs where the block grades deviate

December 2019 87

are generally associated with areas of low data. Figure 37 and Figure 38 are swath plots for copper and gold respectively for domain 1203 by elevation. Both figures show the OK grades corresponding very well, particularly through the areas of higher data density as indicated by the count bars.

Figure 37 - Domain 1203 Swath Plot - Copper

Figure 38 - Domain 1203 Swath Plot - Gold

0

500

1000

1500

2000

2500

3000

3500

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

450

420

390

360

330

300

270

240

210

180

150

120 9

0

60

30 0

-30

-60

-90

-12

0

-15

0

-18

0

-21

0

-24

0

-27

0

-30

0

-33

0

-36

0

-39

0

-42

0

BLO

CK

S

CU

(%

)

LEVEL (M)

Cu(%) 1203 Domain

Blocks Krig 1203 NN 1203

0

500

1000

1500

2000

2500

3000

3500

0

0.2

0.4

0.6

0.8

1

1.2

1.4

450

420

390

360

330

300

270

240

210

180

150

120 9

0

60

30 0

-30

-60

-90

-12

0

-15

0

-18

0

-21

0

-24

0

-27

0

-30

0

-33

0

-36

0

-39

0

-42

0

BLO

CK

S

AU

(PP

M)

LEVEL (M)

Au(ppm) 1203 Domain

Blocks Krig 1203 NN 1203

December 2019 88

14.7. Resource Model Pit Optimization Mineral Resources exclusive of Mineral Reserves that are amenable to open pit mining methods at Salobo represent sulphide mineralization that is adjacent to the current Mineral Reserve pit plus Inferred Mineral Resources within the Mineral Reserve pit. There are no oxide Mineral Resources. The Mineral Resource estimates were prepared by Vale staff using the following design approach:

• Prices of US$3.18/lb for copper and US$1,290/oz. for gold for Whittle optimization of the Mineral Resource pit.

• Determine reasonableness of Mineral Resource pit extents, such as impact on planned mine infrastructure (waste lay down areas, processing facilities); distribution of deleterious mineral; adequateness of current waste storage capacity.

• Consider a cut-off grade of 0.253% copper equivalent (CuEq), consistent with Salobo’s cut-off grade used for Mineral Reserves reported in 2019.

• Prepare a provisional extension of the LoMP production schedule to include material above this cut-off grade.

• Apply metal price and exchange rate assumptions to forecast cash flows, including appropriate provision for sustaining capital and operating costs.

• Scheduling waste and mineralized material.

• Determine if the Mineral Resource estimates demonstrate a positive cash flow. External mining dilution and mine loss were not applied. Table 26 summarizes the technical and economic parameters used for optimizing the Mineral Resource pit. Figure 39 is an isometric view of the resulting resource pit. The Mineral Resource pit bottoms out on the bottom of the block model. Deep drilling has confirmed mineralization below the base of the current block model and it is anticipated the next block model update in 2020 will extend the base of the block model deeper.

Table 26 - Mineral Resource Open Pit Optimization Assumptions

Parameter Unit Details

Copper sale price $US/lb. $3.18 Gold sale price $US/oz $1,290 Exchange rate BRL/$US 3.8 Mining method Open Pit Cut-off %CuEq. 0.253 Mineability % 100% Dilution % 4% (5 yr) 3% (LoMP) Mine production rate – ore M tonnes/year 40 Mine production rate – waste M tonnes/year 86 Mine full operating cost $/tonne mined 3.54 Mine sustaining capital cost $/tonne mined 0.5 Overall processing cost $/tonne ore 10.15 Site G&A M $US/year 25.9 Overall processing Cu recovery % (-2.5362*(1/Cu))+90.674 Overall processing Au recovery % (1.0173*RecCu)-20.357

December 2019 89

Figure 39 – Mineral Resource Pit – Isometric View NNE

14.8. Classification of Mineral Resources Mineral Resource model blocks are classified as Measured, Indicated or Inferred Mineral Resources, in accordance with CIM guidelines (CIM, 2014). Vale’s long-term mine planning and design process then converts Measured and Indicated Mineral Resources within the current LoMP open pit design into Proven and Probable Mineral Reserves, respectively. Therefore, Mineral Resources at Salobo Operations are stated exclusive of Mineral Reserves. Mineral Resources at Salobo Operations thus comprise:

• Measured, Indicated and Inferred Mineral Resources outside, but adjacent to, the current LoMP open pit design, that are considered to have reasonable prospects for eventual economic extraction based upon it’s analysis of an optimized pit shell as described above.

• Inferred Mineral Resources located within the current LoMP open pit design that are considered too speculative geologically to have the economic considerations applied to them that would enable them to be categorized as Mineral Reserves. In the LoMP open pit production schedule, this Inferred Resource material is planned as waste material. If during the mining process this material is confirmed as ore as per blast holes analysis results, this would be defined as ore for the short-term grade control purposes.

December 2019 90

14.9. Mineral Resource Tabulation Mineral resource estimates (Table 14) represent in-situ tonnages and grades that take into account the minimum block size that can be selectively extracted. Mining recovery has not been applied to the Mineral Resource estimates but a factor for mining dilution is included in the reserve estimation. The Mineral Resources estimated in this statement are consistent with the requirements outlined in NI 43-101 and the Vale Base Metals 2019 Guidelines and Standards for MRMR Reporting. A comparison of the Mineral Resources from 2019 to 2018 is presented in Table 27 and summary of the changes in Table 28. For Measured and Indicated Resources, there was a 6% decrease in tonnes, largely due to the conversion to Proven and Probable Mineral Reserves. Inferred Resource tonnes increased by 3% as a result of downgrading a small portion of Measured and Indicated Resources.

Table 27 - Mineral Resources from 2018 to 2019

2018 2019

% Diff

% Diff

Mea Ind M&I Inf Mea Ind M&I Inf M&I Inf

Within Current

LoMP Open Pit

M Tonnes - - - 33.0 - - - 21.9 -

-34%

Cu % - - - 0.50 - - - 0.50 - 0%

Au (g/t) - - - 0.24 - - - 0.20 - -

17%

Adjacent to Current

LoMP Open Pit

M Tonnes 32.8 172.3 205.1 138.2 1.2 192.3 193.5 154.1 -6% 12%

Cu % 0.73 0.63 0.64 0.6 0.67 0.61 0.61 0.60 -5% 0%

Au (g/t) 0.43 0.31 0.33 0.3 0.42 0.31 0.31 0.30 -6% 0%

Total

M Tonnes 32.8 172.3 205.1 171.2 1.2 192.3 193.5 176.1 -6% 3%

Cu % 0.73 0.63 0.64 0.60 0.67 0.61 0.61 0.59 -5% 0%

Au (g/t) 0.43 0.31 0.33 0.30 0.42 0.31 0.31 0.29 -6% 0%

December 2019 91

Table 28 - Changes to Mineral Resources from 2018 to 2019

M Tonnes Cu (%) Au (g/t)

2018 Measured Mineral Resources 32.8 0.73 0.43

Less mining (includes forecast to year end)

Less Mineral Resources converted to Proven Mineral Reserves -12 0.73 0.59

Less conversion to Indicated Mineral Resources -14.9 0.39 0.2

Less conversion to Inferred Mineral Resources -4.8 1.8 0.72

Less downgrade to exploration target

Less stockpile reclaim

Less sterilization

Re-evaluation

Plus Measured Mineral Resources reclassified from Mineral Reserves

Plus upgrade from Inferred Mineral Resources

Plus new Measured Mineral Resources from drilling

2019 Measured Mineral Resources 1.2 0.67 0.42

2018 Indicated Mineral Resources 172.3 0.63 0.31


Less conversion to Proven Mineral Reserves

Less conversion to Probable Mineral Reserves

Less upgrade to Measured Mineral Resources

Less conversion to Inferred Mineral Resources


Less sterilization

Re-evaluation 5.1 0.71 0.42

Plus reclassified from Mineral Reserves

Plus re-categorized from Measured Mineral Resources 14.9 0.39 0.2

Plus upgrade to Indicated from Inferred Mineral Resources

Plus new Indicated Mineral Resources from drilling

2019 Indicated Mineral Resources 192.3 0.61 0.31

2018 Inferred Mineral Resources 171.2 0.60 0.30


Less mineral conversion to Proven Mineral Reserves

Less conversion to Probable Mineral Reserves

Less upgraded to Measured Mineral Resources

Less upgraded to Indicated Mineral Resources

Less conversion to exploration target (ET)

Less sterilization

Less re-evaluation

Plus re-categorized from M&I Mineral Resources 4.8 1.8 0.7

Plus new Mineral Resources from drilling (upgrade from ET)

2019 Inferred Mineral Resources 176.1 0.59 0.29

14.10. Comments on Section 14 In the opinion of the QPs, the Mineral Resources have been estimated and classified in accordance with 2014 CIM Definition Standards. To the extent known to the QPs, there are no known environmental, permitting, legal, title related, taxation, socio-political or marketing issues that could materially affect the Mineral Resource estimation that are not documented in this Report.

December 2019 92

As detailed in Section 15.7, reconciliation work has shown that the block model estimates are accurately predicting production.

15. Mineral Reserve Estimates

15.1. Mineability and Dilution Dilution and mineability have both been applied in the conversion of Mineral Resources to Mineral Reserves. Mining dilution is the unavoidable mixture of different quality materials during the operation and can be split in two forms: planned (or internal) mining dilution and unplanned (or external) mining dilution. Mining recovery is the failure to recover all predicted ore in the operation, which can be the result of mining dilution but also of several other factors. Internal or planned dilution represents zones of mineralization below the cut-off grade that are unavoidably mined along with the mineralization above the cut-off grade due to the selectivity of the Selective Mining Unit (SMU). It is related to the production rate, equipment size, and, in open pit operations, to the size of the average-grade-polygons delivered to the operations to be mined. Planned dilution is included in the Mineral Resource tonnage and grade through the regularizing of the block model to the SMU size. External or unplanned dilution is the result of introducing different quality material within the planned ore via blasting and equipment operation. External dilution has been incorporated by increasing tonnes by 4% with zero grade in the 5-year plan and thereafter by 3%. The reduction to 3% is expected to be achieved through incorporating blast movement modeling into the material boundaries. Mining recovery is a factor that decreases the ore tonnage predicted by the in-situ geological model, by failing to extract it as ore. In 2019, an average mining recovery factor of 100% in the conversion of Mineral Resource tonnage and grade to Mineral Reserve tonnage and grade was used.

15.2. Pit Optimization The last pit optimization update was executed in 2016 and optimized shells were generated using Whittle Four-XTMv4.4. software and metal prices, recoveries, geotechnical information, and costs as described in the following sections. After detailed pit optimization the revenue factor 1.0 shell was chosen to guide the redesign of the ultimate pit.

December 2019 93

15.2.1. General Assumptions The topographic surface was adjusted to account for the expected mining of the open pit phases and construction of waste dumps to December 31, 2019. Optimized pits were not constrained by any surface infrastructure as it is all located beyond the economic pit limit. A discount rate of 8% and an assumed mining decent rate of five benches per year were applied during the optimization process.

15.2.2. Mineral Resource Model The Geology Department created an updated Mineral Resource model in 2019 as described in Section 14.6. The Mineral Resource model is a partial model that calculates the portion of the blocks within the low and high-grade domains. This partial model was then regularized to a whole block model. The whole block model uses a 30 m x 30 m x 15 m block size which adequately represents the amount of internal dilution currently experienced in mining the Salobo deposit. The other modifying factor used is the incorporation of unplanned mining dilution in the mining plan scheduling. Mining recovery of 100% is assumed during the pit optimization.

15.2.3. Mining Costs For pit optimization the base mining costs at the 250 bench is $3.54/t mined for fresh rock and $3.20/t mined for saprolite, with an average overall mining cost of $3.37/t. The mining unit costs were increased by $0.00412/t for each 15 m bench above the 250 bench and increased by $0.0440/t for every bench below the 250 bench.

15.2.4. Processing Costs The process cost applied in pit optimization is a constant $7.91/t milled, including the processing sustaining costs. The processing cost does not include any incremental drilling, blasting, loading, hauling, or ore control costs. This has the effect of not increasing the marginal cut-off grade. The processing cost does not include any tailings expansion or other sustaining capital costs.

15.2.5. Recovery Copper and gold recovery are estimated based the empirical equations shown below:

𝐶𝑜𝑝𝑝𝑒𝑟 𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦 = 90.674% −2.5362%

𝐶𝑜𝑝𝑝𝑒𝑟 𝐺𝑟𝑎𝑑𝑒

December 2019 94

𝐺𝑜𝑙𝑑 𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦 = (1.0173 𝑥 𝐶𝑜𝑝𝑝𝑒𝑟 𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦) − 20.357 To calculate payable metal average payable factors of 96.7% for copper and 93.94% for gold were applied to the recovered metal.

15.2.6. Over Head Costs General and administrative operating costs were modeled at a constant rate of $1.60/t milled.

15.2.7. Refining, Freight, and Royalties Concentrate was modeled with a copper grade of 38%, 9.5% moisture, and a 0.5% loss in transit. Refining costs inclusive of any penalties are modeled as a cost of $0.57 / lb of copper and $0.52 / oz of gold. Freight consists of road transport to Parauapebas, handling and load out at Parauapebas, rail transport to São Luis, handling and load out at São Luis, and overseas transport. The total freight cost is $0.103 / lb copper or $77.66 / wet tonne of concentrate. There are no 3rd party royalties that are applicable to the property.

15.2.8. Sustaining Capital Sustaining capital was divided between the mine, process plant, and G&A. The sustaining capital was based on the estimated capital requirements to execute the life of mine plan. The sustaining capital for the mine is $0.56/t mined, the sustaining capital for the mill is $0.53/t milled, and $0.02/t milled for the G&A. The sustaining capital was added to the mining, processing, and G&A operating costs for the Whittle optimizations.

15.2.9. Geotechnical Assumptions The overall wall slopes used in the pit optimization are based on the sectors shown in Figure 40 but are flatter then shown in Table 29 as they include an allowance for the access ramps. The overall angles used range from 48 to 52 degrees.

15.3. Selective Mining Unit The SMU was updated in 2016 to represent the actual selectivity achieved by the mine. The original SMU of 15 m x 15 m x 15 m was increased to 30 m x 30 m x 15 m. The average blast hole pattern in ore is 4.5 m x 4.5 m, resulting in 44 holes per block. This provides good coverage of blast hole samples for grade control purposes.

December 2019 95

15.4. Geotechnical Considerations Salobo has been actively mining for over nine years and to date there has not been any significant wall failures. The pit walls have been monitored continually since 2014 by interferometry radar. When some movement has been detected preventative measures have been successfully implemented to avoid risks to the operators. To minimize the risks associated with rock spilling down the temporary wall between a higher and lower phase, mining is sequenced to avoid one active mining area directly above another. Where this isn’t practical the catch berms in the temporary wall have been increased to provide additional catchment. Geotechnical inspections and monitoring are constantly made for both short and long-range mine plans and are checked against the existing models. Some geotechnical points of attention have been identified by this work and preventive actions were suggested to avoid risks to the operations. Most of them are related to bench cleaning due to rock spillage from blasting movements. A surface drainage system to avoid erosion was installed. The wall designs are shown in Table 29 and Figure 40.

Table 29 – Geotechnical Design Sectors for Salobo Mine

Sector Vertical Berm Spacing Face Angle Berm Width Inter Ramp Angle (m) (degrees) (m) (degrees)

Saprolite and semi-weathered material 7.5 or 15 52 10.0 35

Sector I

I 15 70 8.0 48.1

IA 15 70 8.0 48.1

IB 30 70 12.5 52

IC 30 70 12.5 52

Sector II

II 30 70 12.5 52

IIA 15 70 8.0 48.1

IIB 30 70 12.5 52

Sector III III 30 70 12.5 52

Sector IV IV 30 70 12.5 52

Sector V V 30 70 12.5 52

VA 15 70 8.0 48.1

Sector VI

VI 30 70 12.5 52

VIA

15 70 8.0 48.1

VIB

30 70 12.5 52

VIC

30 70 12.5 52

VID

15 70 8.0 48.1

Sector VII VII 30 70 12.5 52

December 2019 96

Figure 40 – Geotechnical Design Sectors

A review of the geomechanical model of the Salobo Pit was completed in 2019 by Walm Engenharia. The construction of the waste dump and stockpiles are regularly monitored with particular attention to the placement of weathered rock and soil. Geotechnical inspections and monitoring of the Salobo dam are being completed according to the new governmental agency regulation (DNPM Act 70.389/17). The Emergency Action Plan and the periodic dam safety review were concluded in accordance with the current legislation. A radar system monitors the dam and a seismographic station has been installed to record seismic events in Salobo and will be integrated into the national seismic monitoring network. External consultant companies (BVP e WALM) demonstrated the stability of the dam and internal reports are being sent to the Regulation Agency regularly, according to required legislation. The heightening of Salobo dam was divided in 3 stages. The first was completed to the 220 m elevation. The second stage was divided into three phases with the first phase up to the 235 m elevation, the second phase up the 245 m elevation and the third phase up to the 255 m elevation and completed in 2019. There is still 30 m remaining to raise in the third stage to satisfy the LoMP requirements. The third stage, elevation 285 m, will be completed in 2025.

December 2019 97

15.5. Cut-off Grade The marginal cut-off grade (GCu) is applied to a copper equivalent grade (CuEq). The calculation is based on costs, prices and recoveries in the plant. The general expression for the equivalent copper grade and the marginal cut-off grade for defining ore and waste are shown below using the parameters shown in Table 30.

𝐶𝑢𝐸𝑞 = 𝐶𝑢 + (𝐴𝑢 × (𝑃𝑟𝐴𝑢 − 𝐶𝑣𝐴𝑢) × 𝑅𝐶𝐴𝑢 × 𝑅𝐹𝐴𝑢

(𝑃𝑟𝐶𝑢 − 𝐶𝑣𝐶𝑢) × 𝑅𝐶𝐶𝑢 × 𝑅𝐹𝐶𝑢) ×

31.103

2,204.62

𝐺𝐶𝑢 = 𝐶𝑃 + 𝐶𝐺&𝐴 + 𝐶𝐴𝑀

(𝑃𝑟𝐶𝑢 − 𝐶𝑣𝐶𝑢) × 𝑅𝐶𝐶𝑢 × 𝑅𝐹𝐶𝑢 × (100 − 𝑃𝑇) × 2,204.62

Table 30 – Cut-off Calculation Parameters

Parameter Description Value

Cu Copper Grade (%) Variable Au Gold Grade (g/t) Variable

PrCu Copper Price ($/lb.) $3.18 PrAu Gold Price ($/oz.) $1,290

CvCu Copper Selling Cost ($/lb) $0.675 CvAu Gold Selling Cost ($/oz.) $0.52

RcCu Copper Flotation Recovery (%) Calculated RcAu Gold Flotation Recovery (%) Calculated

RfCu Copper Smelting Recovery (%) 96.70% RfAu Gold Smelting Recovery (%) 93.94%

CP Total Cost of Processing($/t) $10.15 CG&A Total Cost of G&A ($/t) $1.95 CAM Additional Cost of Mining Ore ($/t) $0

PT Loss of Concentrate in Transport 0.50%

31.103 Conversion Factor troy oz. to gram 31.13 2,204.62 Conversion Factor tonne to lb. 2,204.62

A marginal cut-off grade of 0.253% was used for copper and is applied in equivalent amounts of copper. This cut-off is used for the determination of Mineral Reserves and mining sequencing. At the Salobo Operations, a strategy to stockpile lower grade ore was adopted to maximize the NPV of the mine which results in an elevated cut-off being used until 2044. Currently 163 M tonnes of ore is stored in the Low Grade Stockpile.

15.6. Mineral Reserve Tabulation The long-term mine planning, pit optimization and design process converts the open pit measured and indicated Mineral Resources into proven and probable Mineral Reserves. All Inferred material contained within the ultimate pit is treated as waste. The Vale 2019 block model forms the basis of Salobo’s Mineral Reserves and Mineral Resources. Mineral reserve estimates are derived from this block model by applying the appropriate technical and economic parameters, within the 2017 ultimate pit design. Key parameters are calculated separately for each discrete mining block, based on geometry

December 2019 98

and mining method, as detailed in this section of the technical report. Copper equivalent grades were calculated using metal prices of $3.18/lb for copper and $1,290/oz for gold. The cut-off grade of 0.253% CuEq applied to the 2019 block model reflects Vale’s forecasts of direct operating costs, recoveries and metal prices, etc. The Mineral Reserve includes planned mining dilution of between 3% and 4% with 100% recovery. Table 31 details the 2019 Mineral Reserves for the Salobo Operations.

Table 31 - December 31, 2019 Mineral Reserves

Grades Contained Metal Classification M Tonnes Cu % Au g/t Cu (M lb) Au (M oz)

Proven 152.7 0.69 0.39 2,319 1.9 Stockpiles (Proven) 163.4 0.45 0.22 1,621 1.2

Probable 832.4 0.62 0.32 11,377 8.6

P&P 1,148.4 0.60 0.32 15,318 11.6 Notes:

1. Mineral Reserve estimates were prepared by Mr. Wellington F. de Paula, an employee of Vale. The Qualified Person for the

Mineral Reserve estimates is Mr. Marcos Dias Alvim, P.Geo., FAusIMM(CP), Long Term Planning Manager, South Atlantic

Operations, Vale Base Metals.

2. Mineral Reserves are classified as Proven and Probable Mineral Reserves based on the 2014 CIM Definition Standards. 3. Mineral Resources are reported above a copper equivalent cut-off of 0.253%, assuming $1,290 per ounce gold and $3.18 per

pound copper. 4. Tonnages are rounded to the nearest 100,000 tonnes and grades are rounded to two decimal places. 5. Contained copper is reported as Imperial pound units and contained gold as troy ounces.

A comparison of the 2018 to 2019 Mineral Reserves estimates is provided in Table 32 and the changes between the 2018 and 2019 Mineral Reserves are shown in Table 33.

Table 32 - Summary of Mineral Reserves from 2018 to 2019

2018 2019 % Diff

Prov Prob P&P Prov Prob P&P P&P

Salobo Pit

M Tonnes 478.2 537.7 1015.9 152.7 832.4 985.0 -3.0%

Cu % 0.69 0.58 0.63 0.69 0.62 0.63 0.0%

Au (g/t) 0.38 0.29 0.33 0.39 0.32 0.33 0.0%

Stockpiles

M Tonnes 141.0 - 141.0 163.4 - 163.4 14.7%

Cu % 0.43 - 0.43 0.45 - 0.45 4.5%

Au (g/t) 0.21 - 0.21 0.22 - 0.22 4.7%

Total

M Tonnes 619.2 537.7 1,156.9 316.1 832.4 1,148.4 -0.7%

Cu % 0.63 0.58 0.61 0.56 0.62 0.60 -1.7%

Au (g/t) 0.34 0.29 0.32 0.30 0.32 0.32 0.0%

December 2019 99

Table 33 - Changes to Mineral Reserves from 2018 to 2019

Tonnes (Million) %Cu g/tAu

2018 Proven Mineral Reserves 619.2 0.63 0.34

Less mining (includes forecast to year end) -33.4 0.75 0.44

Less re-evaluation

Less re-categorize to probable mineral reserves -292.2 0.68 0.37

Less reclassification to Measured Mineral Resources


Less stockpile reclaims

Plus re-evaluation

Plus upgrade probable to Proven Mineral Reserves

Plus Measured Mineral Resources converted to Mineral Reserves

Plus new Mineral Reserves

Plus stockpile additions 22.4 0.54 0.27

2019 Proven Mineral Reserves 316.1 0.56 0.30

2018 Probable Mineral Reserves 537.7 0.58 0.29

Less mining (includes forecast to year end) -22.1 0.55 0.30

Less re-evaluation probable mineral reserves

Less upgrade of probable to proven mineral reserves

Less reclassification to indicated mineral resources


Less stockpile reclaims

Plus re-evaluation 12.6 0.42 0.42

Plus re-categorize proven to probable mineral reserves 292.2 0.68 0.37

Plus Measured Mineral Resources converted to Mineral Reserves 12.0 0.73 0.59

Plus new Mineral Reserves

Plus stockpile additions

2019 Probable Mineral Reserves 832.4 0.62 0.32

2019 Total Mineral Reserves 1,148.4 0.60 0.32

Mining depletions of 55.5 M tonnes occurred during 2019, and account for the main overall reductions in mineral reserve. Of this depletion, 22.4 M tonnes of low grade ore were added back into proven mineral reserve as stockpile material. A recategorization of 292.2 M tonnes of proven mineral reserve to probable mineral reserve occurred because of the recommendations from a drill spacing back analysis study completed in early 2019. There were two small additions to probable mineral reserve; 12.6 M tonnes added through re-evaluation and 12 M tonnes was downgraded from Measured to Indicated Mineral Resource then converted to Probable Mineral Reserve because of changes in resource classification due to the back-analysis study which reduced the quantity of Measured blocks in the resource pit and an increase of Indicated blocks in the reserve pit.

December 2019 100

Table 34 details the Mineral Reserves by mining phases.

Table 34 - 2019 Mineral Reserves Estimate by Phase

Classification Phase M Tonnes Cu (%)

Au (g/t)

Proven & Probable

3 3.8 0.83 0.53

4 101.3 0.70 0.40

5 99.1 0.69 0.33

6 255.2 0.58 0.33

7 205.5 0.62 0.33

8 320.1 0.63 0.31

Stockpiles 163.4 0.45 0.22

Total 1,148.4 0.60 0.32

Figure 41 is a long section through the middle of the orebody depicting the location of Proven and Probable Mineral Reserves within the 2017 reserve pit (red) and the Measured, Indicated and Inferred Resources between the 2017 Mineral Reserve and Mineral Resource (green) pits. The figure also shows the location of Inferred Mineral Resources within the reserve pit at the northwest and southeast ends of the orebody. Future drilling will target the reserve pit Inferred Mineral Resource areas for conversion to Mineral Reserves.

Figure 41 - 2019 MRMR Longitudinal Section (looking southwest)

December 2019 101

15.7. Reconciliation

15.7.1. Method Vale’s reconciliation process at Salobo consists of computing mining call factors that compare tonnage, grade and metal at different measurement points along the process. The key measurement points are: long-range model (LTM), short-range model (STM), polygon model (POL), production (PRD), total ore sent to crusher (TSC), and processed ore (PO). For all measurement points, tonnage, grade and metal are quoted on a dry basis. The following reconciliation factors are calculated at the mine and Figure 42 shows the comparison rationale for each:

• F1 = STM / LTM

• F2 = PRD / STM

• F3 = PO / TSC

• F1’ = STM / LTM (within LTM ore boundaries)

• M1 = POL / STM

• M2 = PRD / POL

• F4 = F1 x F2 x F3

Figure 42 - Production Reconciliation Schematic

December 2019 102

F1 factor

• Compares long-range and short-range models and is performed between topographic surfaces. It allows an assessment of the uncertainty of the geological model boundaries by the evaluation of tonnage differences between the mineralization envelopes.

• Conditional simulations are used in this process to provide a better understanding of grade variability, mining call factors variability and grade estimation related to long-range grade uncertainty.

F2 factor

• Evaluates the mining operation by comparing the predicted tonnage and grade from the short-range model to those measured by the truck and crusher weighing scales.

F3 factor

• Compares the production tonnage and grades with those measured by the plant. F1’ factor

• Compares the LTM and STM but is performed only within the LTM orebody to ensure that the estimation methodology and accuracy related to the production grades are well executed. This factor provides a better understanding of the estimation process controls and parameters, the defined LTM mineralization envelope and the grade variability.

M1 factor

• Compares the short-range model with the polygon model, which consists of the polygons with average grades that are assigned to the operation to be mined. This factor shows what losses may occur due to selectivity by delivering polygons to the operation instead of SMU blocks. It provides a measure of planned dilution.

M2 factor

• Compares the production data measured by the truck and crusher weighing scales to the polygons received by the operation to be mined. This gives an indication of the unplanned dilution and mining recovery occurring within the mining process.

15.7.2. Results Table 35 shows the various reconciliation factors calculated for 2018 and 2019 and descriptions after the table explain the results.

December 2019 103

Table 35 – 2018 / 2019 Calculated Reconciliation Factors

Factor 2018 2019¹

Tonnage Cu (%) Cu (t) Au (g/t) Au (t) Tonnage Cu (%) Cu (t) Au (g/t) Au (t)

F1 1.19 1.01 1.20 1.01 1.20 1.33 0.98 1.31 1.07 1.42

F2 0.90 0.99 0.89 1.01 0.91 0.86 1.02 0.88 1.11 0.96

F3 0.99 0.97 0.96 1.05 1.04 1.02 0.98 1.00 1.01 1.03

F1’ 1.00 1.08 1.08 1.12 1.12 1.00 1.11 1.11 1.30 1.30

M1 0.86 1.02 0.89 1.06 0.92 0.91 1.01 0.92 1.04 0.95

M2 1.04 0.97 1.01 0.96 1.00 0.95 1.02 0.96 1.07 1.01

1As of September 2019

F1 factor

• Shows that the short-range model yields 19% more tonnes than indicated in the long-range model in 2018 and 33% more in 2019. These differences are due to a larger volume of low-grade mineralization in contact with the waste zone that couldn’t be modelled in the long-range orebody. This low-grade mineralization consists of smaller veins that are narrower than the long-range drill spacing can reliably predict.

F2 factor

• Shows that 10% or the modelled ore in the short-range model in 2018 and 14% in 2019 was not mined as ore. This is because the location of the weathered oxide boundary is poorly understood in areas where faults have lowered the oxidation level. Around half of this extra volume of low-grade mineralization modelled in the short-term model is in fact oxide material, which is realized only after blasting. This puts pressure on the short-range planning and scheduling. Methods of improving the short-range interpretation of the oxide surface are currently being investigated including systematic short-range weathering mapping.

F3 factor

• The F3 factors show good control of the ore sent to the crushers. However, the primary crusher weighing scales are not being used and the mass balance calculation is being made using truck weightometers. The installation of an external weighing scale system (with laser scan) is being investigated to confirm truck tonnes and calibrate the truck weightometers.

F1’ factor

• Shows that the average copper and gold grades are higher in the short-range model. Grades higher than predicted in the long-range model were mined in 2019 within Phase 4 of the pit. This was due to wider spaced drilling in that portion of the pit. Copper grades were 8% higher in 2018 and 11% higher in 2019. Gold grades were 12% higher in 2018 and 30% higher in 2019.

December 2019 104

M1 factor

• Shows that the difference between the short-range model and the blasting polygon improved from 14% in 2018 to 9% in 2019.

M2 factor

• Shows a positive difference of 4% between polygons and production in 2018 and a negative 5% in 2019.

Table 36 lists the overall copper production reconciliation between the milled production with the material sent to the crushers. The estimated tonnage and copper grade of material sent to the crusher is derived from estimates based on detailed drilling and grade control mapping. The table shows the improvement in mill vs production reconciliation since 2012.

Table 36 - Mill versus Production

Year Production Mill Mill vs Production

M Tonnes % Cu M Tonnes % Cu Tonnes % Cu Cu tonnes

2014 12.8 0.99 12.5 0.97 -2.0% -2.0% -4.0%

2015 19.4 0.90 20.3 0.88 5.0% -2.0% 4.0%

2016 21.0 0.94 21.4 0.94 2.0% 0.0% 2.0%

2017 23.6 0.95 23.7 0.95 0.4% 0.0% -0.2%

2018 23.9 0.97 23.7 0.95 -1.0% -2.3% -3.6%

20191 16.3 0.97 16.6 0.95 1.8% -1.7% 0.1%

1As of September 2019

15.8. Comments on Section 15 In the opinion of the QPs, Mineral Reserves have been estimated according to 2014 CIM Definition Standards and estimates are based on the most current knowledge, permit status and engineering and operational constraints. Mineral Reserve declaration is supported by a positive cashflow. Improvements in the definition of the oxide / sulfide boundary will improve the tonnage discrepancies observed in the F2 factor and improvements to truck weight estimates will improve the reconciliation between production and plant. Review of the operational pit slope angles through geotechnical examination of the pit wall operation, design of pushbacks, and further geotechnical studies may provide support for steepening of some of the pit walls. The current geomechanical sectors are based on a limited amount of information and this should be further analyzed.

December 2019 105

16. Mining Methods Salobo mine utilizes standard open pit methods, developed in 15 m benches, with trucks and shovels. After drilling and blasting the material, cable shovels, large front-end loaders and hydraulic excavators are used to load this material. A fleet of 240 t and 360 t trucks are used to haul the waste material to waste dumps proximal to the pit or ore material to the primary crusher. Lower grade ore is stockpiled for later processing. The mine planning objective is to mine the ore sequentially in mining phases, considering the largest possible vertical spacing between phases. The plan is to provide an approximately steady annual production of 36.0 million tonnes to the mill. Initial production from the expansion is expected in January 2022. The overall site layout is shown in Figure 51.

16.1. Pit and Phase Designs During 2017, the ultimate pit was redesigned based on the 2016 Whittle pit optimization results and incorporating the revised pit wall designs. The internal phases were also redesigned in 2017 with an eighth phase added to improve the LoMP schedule. The revised phases and ultimate pit are shown in Figure 43.

Figure 43 – Salobo Phases

December 2019 106

16.2. Production Schedule After estimating Mineral Reserves, a practical and executable production schedule is developed by short and long term mine planning teams. The ultimate pit has been subdivided into eight phases two of which have been mined out the remaining six phases form the basis of the life of mine plan. The Mineral Reserves for the remaining phases are shown in Table 37.

Table 37 - December 31, 2019 Mineral Reserve Estimate by Phase

Classification Phase M Tonnes Cu (%)

Au (g/t)

Proven & Probable

3 3.8 0.83 0.53

4 101.3 0.70 0.40

5 99.1 0.69 0.33

6 255.2 0.58 0.33

7 205.5 0.62 0.33

8 320.1 0.63 0.31

Stockpiles 163.4 0.45 0.22

Total 1,148.4 0.60 0.32

In general, the phases have been sequentially scheduled with a maximum ore plus waste production rate of 126 Mtpa feeding 36 Mtpa of ore to the processing plant. This plan achieves the 36 Mtpa plant throughput by processing a portion of the material that would have been stockpiled in the previous 24 Mtpa production plan. Maintaining the high-grade strategy used in the 24 Mtpa plan would require increasing mine production by purchasing additional equipment and hiring personnel which are not currently planned but may be considered in the future. The initial 5 years of the mine plan is shown in Table 38 and the entire schedule is shown in Figure 44.

Table 38 - 5 Year Plan

Period 2020 2021 2022 2023 2024

Ore (kt) 53,437 51,746 43,470 50,059 30,762 Waste (kt) 64,185 65,908 74,174 64,315 75,230

Mine Production (kt) 117,622 117,622 117,622 117,622 117,622 Other Movement (kt) 22,378 22,346 22,355 21,626 20,008

Total Movement (kt) 140,000 140,000 140,000 136,000 126,000

Stripping ratio 1.2 1.3 1.7 1.3 2.5

Plant feed (kt)¹ 23,852 23,847 33,131 35,873 36,000 Plant feed (Cu %) 0.97 0.96 0.85 0.83 0.69 Plant feed (Au g/t) 0.63 0.58 0.45 0.45 0.37

Notes: ¹ Dry metric tonnes; Source: Five-year plan production.

December 2019 107

Figure 44 - Life of Mine Plan (Mineral Reserves Only)

The open pit mine life is approximately 25 years, ending in 2044. However, the process plant will continue to operate by reclaiming stockpiled material until 2052. Phasing of the open pit development and application of the cut-off grade strategy allows higher grade ore (above 0.90% Cu) to be processed in the initial years of the operation. Once the stockpile has been reclaimed, there are additional Mineral Resources in the Salobo pit that could sustain the operation for another 10 years.

16.3. Grade Control Grade control at Salobo Mine uses samples of drill cuttings collected from blastholes and applies procedures developed at the nearby Vale Sossego mine. All ore blastholes are sampled, and the grade control geologist determines which waste blastholes are sampled to ensure mineralization matches the interpretation in the geological model. Surveyors measure drill hole collar locations using high precision global positioning system (GPS) equipment. As described in Section 11.1.2, samples of drill cuttings are collected from the entire length of the blasthole, including the subdrill. The sample is homogenized and reduced to 2 kg using a Jones splitter before it is bagged and numbered prior to dispatch to the analytical laboratory.

December 2019 108

Blasthole samples are analyzed at the Salobo laboratory and the results are forwarded to the geology department to be entered into the ore control database. Ore polygons are defined, based on the assay results and taking account of where the blasted material was thrown. This information is uploaded to the GPS units of the operating shovels and loaders to guide the mucking operations. A dispatch system is used to control the activities of all mine equipment, and compliance to the mine plan is monitored on a monthly basis. Adherence to the mine plan is recognized as key to achieving the overall production forecast.

16.4. Mining Equipment The Salobo bulk mining operations primarily utilize large electric (rope) shovels for ore and waste production. Hydraulic shovels are used for the oxide saprolite and transition material where a lower ground pressure is required. Wheel loaders are used for miscellaneous clean up jobs and for backup of the shovels when needed. A fleet of off-road haul trucks are used to transport the material to either the waste dump or the primary crusher stockpiles. Low and medium grade ore is stockpiled near the open pit. Cycle times for haulage calculations are determined for each mining period using the Mine Haul software. The track dozers are assigned to maintain the production areas, waste dumps and cleaning up the benches. Wheeled dozers, road graders and water trucks complete the remainder of the auxiliary equipment fleet. Table 39 provides a summary of the mining fleet at the Salobo Operations. The equipment listed is used to develop, drill, blast/muck and haul ore from the active mining levels. The equipment consists of electrical and diesel powered drills of 12 1/4'”, 10” and 6 1/2” diameters, cable shovels 42 yd3

and 63 yd3, hydraulic excavators 38 yd3 and wheel

loaders of 33 yd3 capacity. The diesel drills and the wheel loaders are mainly used for ore

exploitation that needs more mobility, and electrical drills and cable shovels are used for waste removal and bulk ore portions. The wheel loaders support the larger units in narrow areas and in opening new accesses. Komatsu 830E-AC (240t), Caterpillar 793 (240 t) and 797 (360 t) trucks were selected to haul all material from the pit. The auxiliary units are bulldozers (D475A-2, Cat D11, D375A, Cat D10 and D6R); this equipment is necessary to maintain the production areas, waste rock dump, and cleaning the material on the benches. Wheeled tractors, motor graders and water trucks complete the auxiliary equipment fleet. Table 39 details the current fleet.

December 2019 109

Table 39 - Actual Production Mining Fleet

Fleet Qty.

Avg. Loading 12

BE 495 HD (42 yd3) 4 BE 495 HR (63 yd3) 1 PC 5500 (38 yd3) 4 L 1850 (33 yd3) 3

Avg. Hauling 51

Kom 830 (240 t) 18 Kom 930 (320 t) 3 CAT 793 (240 t) 14 CAT 797 (360 t) 16

Avg. Drilling 21

Pit Vipper 351 (12 1/4") 6 BE 49 HR (12 1/4") 6 Pit Vipper 235 (9 7/8") 3 ROC-L8 (6 3/4") 5 ROC-D7 (5 1/2") 1

16.5. Manpower The mine operates on a continuous schedule with three shifts per day of 8 hours each. Approximately 10 days per year are planned as lost production delays due to poor weather conditions (i.e., rain and fog). Forecast mine manpower utilization takes into account delays for training, blast moves and other operational delays.

16.6. Ore Stockpiles & Waste Disposal Low-medium grade ore and waste rock from the mine are stored in three locations along the perimeter of the pit as shown in Figure 43. The main waste rock dump is to the west of the pit and contains both oxidized and fresh rock. Geotechnical investigations were conducted to develop the dump design parameters. A 35% swell factor is used to compute the required storage volumes in the stockpiles and waste dumps. Material is end-dumped in 20 m high lifts with 10 m berms between lifts. The bench face angles range from 32 to 35 degrees, depending on the angle of repose for the material (see Table 40). Including the berms, the overall slope of the dumps ranges from 2H:1V to 2.5H:1V. The resulting slopes were shown to have an estimated factor of safety of 1.5 against large scale circular slip failures.

Table 40 - Waste Dump & Stockpile Design Parameters

Units Waste Dump Ore Stockpile

Lift Height m 20 10 Angle of Repose degrees 32-35 34 Berm Width m 10 15 Overall Slope degrees 20 25

The waste materials and the low-medium grade ore have been characterized as having low acid rock drainage potential (Brandt, 2003). However, there is a concern with fluorine

December 2019 110

leaching from the finer grained oxidized materials (saprolite). Accordingly, mineralized saprolite material is encapsulated within the waste rock dump to control infiltration of surface water and minimize resultant leaching. The long-term storage of the medium and low grade material in a tropical environment may lead to some oxidation of contained sulphide minerals, impacting recovery of metals during eventual processing of the stockpiles.

16.7. Mine Services Water management (including pit dewatering, and control of runoff within the open pit and surrounding area) requires additional attention during the rainy season. With over 1.9 m of rainfall each year, sumps and pumps need to be well managed to maintain the roads and pit working surfaces. The Operation recognizes this and has allocated appropriate resources to this task. During the dry season, dust control is maintained through the use of water trucks. Evaluations are on-going to determine the effectiveness of additives, such as calcium chloride for dust control.

16.8. Comments on Section 16 In the opinion of the QPs, the mining methods, equipment, overall design and the production rate assumptions used to develop the LoMP and Mineral Reserves are reasonable and achievable. A technical analyses of Mineral Resources below the Mineral Reserve pit should be studied to determine if the low-grade stockpile reclaiming in the later years of operation can be delayed.

17. Recovery Methods

17.1. Process Flowsheet The process flowsheet has evolved through the various study phases of the Project, incorporating the additional knowledge gained from metallurgical testwork and the relative importance of the identified lithologies in the Mineral Resource and Mineral Reserve estimates. In particular, the following stages of Project development contributed to the evolution of the retained flowsheet.

• The CVRD and Anglo American testwork program, from 1986–1987, provided the basis for a prefeasibility study completed by Bechtel in 1988. At this stage, fluorine contamination of the concentrate was recognized

December 2019 111

• The SMSA testwork program, culminating in a pilot plant campaign at the CRC, performed between 1993 and 1998, provided additional data for a final feasibility study completed by Bechtel

• Locked-cycle flotation tests, flotation variability, and grinding studies, completed in 2003 and 2004, were used by Fluor Daniel to complete a second feasibility study in 2004, which evaluated production scenarios at 12 Mtpa and 24 Mtpa.

• A trade-off study using high-pressure grinding rolls (HPGR) for tertiary crushing as an alternative to conventional semi-autogenous grinding (SAG), conducted from 2005–2006. The data thus collected were used by Kvaerner to prepare a trade-off study, from which the HPGR approach was adopted.

HPGR were retained instead of SAG mills because of the high magnetite (and copper) content of critical-size pebbles that would have been removed with the magnet protecting the pebble crushers, and therefore requiring additional re-handling (per Vale’s experience at Sossego). In addition, the relatively high ore hardness and its expected variability as different mixtures of ore lithologies are introduced as plant feed, would have caused high-frequency variability in plant throughput in a typical SAG mill–ball mill–pebble crusher (SABC) circuit. Phase I of the Salobo plant (Salobo I) was designed to process 12 Mtpa of ore, to produce approximately 100 kt of copper-in-concentrate annually. Production commenced in June, 2012. The Salobo II plant permitted a doubling of the nominal plant throughput, to 24 Mtpa, with an annualized copper-in-concentrate production of approximately 200 kt. The Salobo II plant was commissioned in June 2014 and is basically a mirror-image of Salobo I, i.e. essentially two identical, parallel, production lines. Salobo I was designed to operate 365 days per year, 24 hours per day and with a targeted 90% of actual operating time, accounting for availability and utilization. Salobo II started operation in June, 2014 and is designed for a targeted up-time of 90%. The overall simplified process flow diagram is illustrated in Figure 45. Apart from the inclusion of HPGR for tertiary crushing duty, ahead of ball milling, the circuit is conventional, but with the flotation cleaning circuit making extensive use of flotation columns, to reduce entrainment of F-bearing non-sulphide gangue minerals such as fluorite and biotite. The existing processing plants, line 1 and line 2 (Salobo I and II), have proven capable of processing nominally 12 Mt of ore annually each. Vale have decided to increase the ore processing capacity to a total of 36 Mtpa by the implementation of the Salobo III Project.

December 2019 112

17.2. Plant Design

17.2.1. Salobo I and II The whole plant is extensively instrumented. All signals are provided to a distributed control system (DCS), allowing for the remote activation and stoppage of equipment, as well as the monitoring of the status of process equipment and of the metallurgical performance of the plant. A manned control room is used to implement changes to the circuit, with the instructions relayed from floor supervisors via radio. Run-of-mine ore at 2.5 m top size is hauled in 240 t trucks and crushed in one of two 60” x 89” primary gyratory crusher (600 kW motor), rated for 1,826 t/h each, to a product size distribution with 80% of the mass passing 152 mm while operated with an open-side setting (OSS) of 140 mm. The dump pocket capacity is equivalent to the volume of 2.5 trucks. Primary crushed ore is conveyed to a common crushed ore stockpile which has a live capacity of approximately 24,800 t and a total capacity of 73,400 t. Four coarse ore stockpile reclaim feeders are used to feed onto the primary screen feed conveyor which feeds two operating double-deck vibrating screens. The screens have a 100 mm aperture top deck and 55 mm aperture bottom deck to yield and underflow product sizing of 80% passing 38 mm. Screen oversize is crushed in two MP-1000 cone crushers (746 kW motors) in a standard closed circuit. A third screen and crusher were added to the original two units with the Salobo II plant. These units are typically on stand-by. Secondary-crushed product is then conveyed in a 2 km long pipe conveyor running at a speed of 2.5 m/s to the secondary crushed ore stockpile. This stockpile has a total capacity of approximately 171,000 t and a live capacity of about 75,000 t. Two parallel lines of four operating reclaim feeders each are then used to reclaim the crushed ore and deliver it to the High Pressure Grinding Roll (HPGR) circuit via the two stockpile reclaim conveyors merging into a single line of transfer conveyors leading to the HPGR silos feed conveyor, equipped with a shuttle head. This unit delivers ore into one of four concrete silos, providing approximately 20 min of surge at nominal capacity. A reversible feed belt conveyor and feed belt feeders then feed each of the four HPGR units.

December 2019 113

Figure 45 - Simplified Process Flowsheet

Each HPGR unit has a drum 2.0 m diameter by 1.5 m wide. The maximum feed size is 55 mm and the HPGR product is exhibiting 80% passing 17 mm while operating with a 40 mm gap and at 150 bars of hydraulic pressure applied to the floating roll. The crushed HPGR product is discharged via the product collection conveyor and is then screened at 8 mm on the bottom deck of banana screens, with the top deck aperture set at 15 mm. There are a total of eight operating screens, with half dedicated to the HPGR of either Salobo I or Salobo II. The screen undersize, at 80% passing 6 mm, discharges directly into one dedicated ball mill discharge sump. The screen oversize is recirculated back via the screen oversize collection conveyor to the HPGR silos feed conveyor for further crushing. The circulating load is typically 110% around this circuit. Slurry in the ball mill discharge sump is pumped to a battery of ten 660 mm hydrocyclones, of which seven are typically operating. Hydrocyclone underflow is fed by gravity to an overflow ball mill of 7.9 m diameter by 12.2 m long, equipped with a 17 MW gearless motor. There are four ball mills operating in closed circuit, each with a dedicated hydrocyclone cluster. Ball mill discharge feeds into the discharge sump for recirculation to the hydrocyclones. The design grinding circuit product is set at 80% passing 106 μm. Hydrocyclone overflow advances to the Rougher 1 flotation circuit at 45% solids by weight. The ball mills were designed to operate at a 30–35% ball charge using 76 mm diameter steel balls and with a circulating load of approximately 300%. These conditions were adjusted by the operations, now showing use of a 30% ball charge. Under these conditions, 15 MW are drawn from the mill motors. A higher ball charge would reportedly require the addition of a retainer ring at the mill discharge. The circulating load is about 200%. The flotation circuit is of conventional design but the cleaning circuit is making extensive use of column flotation, in order to improve rejection of gangue contaminants carrying fluorine values. Lime is added at the front end of the circuit to raise the pH to about 10.

December 2019 114

Addition of NaHS is made ahead of roughing so as to clean the surfaces of the bornite and increase its recovery. PAX and a dithiophosphate are used as the primary and secondary collectors, respectively. Frothing is provided by propylene glycol and methyl isobutyl carbinol (MIBC). Rougher 1 (e.g. rougher) flotation is carried out in four parallel lines (one for each ball mill) of two cells each. The cells are mechanically agitated units of 200 m3 capacity, providing six minutes of design retention time. The Rougher 1 concentrate advances to the cleaning circuit. The Rougher 1 tailings advance to the Rougher 2 (scavenger) circuit consisting of four lines, with each line containing six mechanically-agitated 200 m3 cells, for a nominal retention time of 39 min. Staged Flotation Reactors (SFR’s) have been installed on the rougher tailings. The concentrate from the SFR’s reports to concentrate regrinding. SFR tailings gravitate to the tailings storage facility (TSF), while the concentrate advances to the regrinding circuit. The cleaning circuit is divided into three upgrading stages and closed by a cleaner–scavenger bank of conventional agitated cells. The arrangement of each upgrading stage is typical, whereas the concentrate of one stage advances to the next one and the tailings are moved back to the previous stage. Exceptions are found with the Cleaner 1 tailings, proceeding to the cleaner–scavenger and Cleaner 3 concentrate, which is the final concentrate. The Cleaner 1 circuit consists of 16 column cells, each 6 m diameter x 14 m height, arranged in four lines of four cells each. Design residence time is 39 min. The Cleaner 1 columns are fitted with a Microcel sparging system, introducing flotation air to recirculated slurry pumped through static mixers. All of the other columns only use more standard air spargers. The concentrate from the Cleaner 1 circuit advances to the Cleaner 2 circuit, consisting of eight cells, in four lines of two columns each, of 4.3 m diameter x 14 m height, for a design retention time of 34 min. Concentrate from the Cleaner 2 circuit advances to the Cleaner 3 circuit, consisting of four cells, in four lines of one cell each, each column 4.3 m diameter x 14 m height for a design retention time of 39 min. The tailings of Cleaner 1 are fed into the cleaner–scavenger section, made of four lines of four 200 m3 agitated cells each. The tailings of this stage join the Rougher 2 tailings to form the complete plant tailings stream, directed by gravity to the TSF. The cleaner–scavenger concentrate is combined with the Rougher 2 concentrate and undergoes regrinding in one of four vertical mills fitted with 1.1 MW motors. These mills, filled with 20 mm diameter steel grinding media, are operated in closed-circuit with one dedicated cyclone cluster per mill, ensuring a regrinding circuit product at 80% passing 20 µm.

December 2019 115

The final concentrate exiting Cleaner 3 is pumped to one of two 15 m diameter high-capacity thickener, producing an underflow slurry at 65% solids. This slurry is transferred to a surge tank ahead of the concentrate filters. The concentrate is dewatered further through the use of four pressure filters, each with a horizontal frame holding 50 plates of 1,500 mm x 1,500 mm. A typical filtration cycle lasts 18 minutes. The filtered concentrate has a residual moisture content of about 11%. It is stockpiled below the filters in a covered concentrate storage area holding 6,000 t. Concentrate is reclaimed by front-end loader and loaded into trucks at a nominal rate of 1,500 wmt/d. The concentrate is weighed to about 27 wmt in the trucks using a static scale and delivered to a rail spur storage area at the town of Parauapebas, some 94 km away. The warehouse can hold 16 kt of concentrate, allowing for blending when required. The concentrate is reclaimed by front-end loader and loaded into 80–90 wmt railcars carrying it to the port of Itaqui, in São Luís, in trains of 100 railcars. The concentrate is stored there in an enclosure with a capacity of 50 kt, while awaiting loading into boats at a rate of 1,100 wmt/h. Sampling of the concentrate is carried out at the Port of Itaqui, in lots of 500 wmt, when the material is reclaimed by loader and placed on the conveyor system feeding it into ships. Shipment weights can vary from 13 kt to 45 kt, with two to three shipments completed per month. The combined flotation circuit tailings (Rougher 2 and cleaner–scavenger tailings) flow by gravity from the plant to the TSF, located directly north of the processing plant. Tailings are dumped from a single-point discharge and create a beach on the south side of the dam. Over the mine life, several phases of dam raising with mine waste will be required to provide the required storage volume. Vertical pumps installed on pontoons pump recycled tailings water back to the process plant, accounting for over 95% of the total process water requirements. A summary of the main process equipment is provided in Table 41 for Salobo I and Salobo II.

17.2.2. Salobo III

Summary The Salobo III Project includes all equipment and unit operations necessary for the processing of copper ore, from the receipt of ROM in primary crushing to the storage of concentrate at the plant, including all utilities, infrastructure and operational and administrative support functions. The process route is very similar to the existing processing plant and basically consists of the stages of comminution (crushing and grinding), classification (wet screening and cycloning), concentration (flotation), regrinding and solid-liquid separation (thickening and filtering).

December 2019 116

At the new plant, the ore will be transported from the mine or temporary stockpile by the mine’s off-road trucks and will pass through the primary crushing stages (gyratory crusher) to a stock-pile, primary screening in a closed circuit with cone crushers and transported through long distance conveyors (TCLD) to the concentration plant. Ore is then recovered from the pile and crushed through high-pressure grinding roll (HPGR), in a closed circuit with secondary wet screening, and ball mill grinding. The ground product will feed the flotation circuit composed of cells, columns and vertical regrinding mills. The concentrate will be thickened and filtered. The tailings will be sent by gravity to the impoundment. An overall mine site layout is shown in Figure 46 and the principal features of the Salobo Mine and Mill Complex are identified along with the planned location of the Salobo III main components in relation to the existing mining facilities and infrastructure.

Figure 46 - Mine site layout with Salobo III

As in the existing facilities, the concentrate, after filtering, will be transferred to the railway terminal in Parauapebas by 35 t trucks. From Parauapebas, the concentrate will travel through Ferrovia to the Port of Itaqui, in São Luis, from where it will be shipped for international trade. The main operations involved in the processing of copper sulphide ore are indicated in the illustrative flow chart (Figure 47).

December 2019 117

Figure 47 - Salobo III Simplified Process Flowsheet

The primary and secondary crushing circuit’s design throughput is 1.29 the design throughput of the HPGR and grinding and flotation plant. This factor corresponds to a 45% additional rate of recovery of the beneficiation pile for HPGR feeding (extra capacity available to accelerate the formation of the beneficiation plant regularization pile). The process route will comprise three major processes, as indicated below:

• Comminution:

o Primary crushing

o Primary screening and secondary crushing

o HPGR size reduction and screening

o Ball mill grinding and classification in hydrocyclones

• Flotation and regrinding:

o Rougher I

o Rougher II

o Staged Flotation Reactor (SFR)

o Cleaner I

o Cleaner II

o Cleaner III

o Cleaner scavenger

December 2019 118

o Regrinding

o Tailings transport and discharge

• Dewatering of concentrate:

o Thickening

o Pipeline transport to filtration facility

o Filtering

The processes are described in the following sections.

Primary Crushing and Coarse Ore Stockpile The ore, transported from the mine by 240/360 t trucks, will be discharged directly on both sides of the feed hopper of the 60”x89” primary gyratory crusher. The hopper's capacity is 800 t, to accommodate two 360 t trucks. Two rock breakers will be installed to process the oversize. The product of the primary crusher, with about 80% passing through 152 mm, will be discharged onto a 2,600 mm wide conveyor for transport to the coarse ore stockpile. The conveyor will be equipped with a weigh scale to record production. The primary crusher's feeding area will be covered, where the 100/25 t capacity crane (which is intended for the maintenance of the crusher). The conical crushed ore storage stockpile will have a useful capacity of 25,400 t (12,700 m³), providing an autonomy of 9 to 14 hours of operation for primary screening and secondary crushing. Dozers can be utilized to provide additional feed to the secondary crushing area. Three plate feeders will be used to reclaim the coarse ore. Two of these feeders are operational and one spare, however all feeders will be capable of operating simultaneously. The feeders will transfer the coarse ore to the belt conveyors which will feed the primary screening and secondary crushing station. A belt scale will be installed to monitor and control the secondary crushing. An electric hoist with a capacity of 7.5 t will be used for maintenance of the drive sets for the plate feeders. Another 7.5 t electric hoist will service the drive assembly of the secondary crushing feed conveyor. A sonic fog dust suppression system will be installed around the feeders within the reclaim tunnel.

December 2019 119

Closed Circuit TV (CCTV) will be installed to monitor discharge and transfer points within the conveying system.

Secondary Crushing The crushed material from the coarse ore stockpile will be discharged onto a series of conveyors which will feed secondary crushing feed bin with a capacity of 600 t (300 m3), providing a 10-minute surge capacity. Belt feeders under the bin will feed primary screening. The two primary vibrating screens (12’ x 24’) are designed to be both operating, with layout provided for a future third unit. The screens will be double deck banana type with 75 mm and 40 mm operating apertures respectively. The plus 40 mm size fraction will be directly discharged to two MP 1000 secondary cone crushers operating in closed circuit. Primary screen undersize at 100% passing 55 mm is final secondary crushing product and is conveyed to the long-distance belt conveying system. The secondary crushers will be protected by metal detectors and magnets and will have a belt scale for process control and monitoring. An overhead crane with a capacity of 60/10 t will serve the screens and crushers. Separate 10 t electric hoists will provide maintenance to the drive assemblies of the various conveyors. Closed-circuit television (CCTV) cameras are provided for the visual monitoring of key points in the process and will allow remote operation through the control room. A baghouse dust collection system will service the secondary crushing and screening area while a sonic fog dust suppression system will be employed on the conveyors.

Long Distance Belt Conveyors The long-distance belt conveyor system (TCLD) will transport the secondary crushed ore to the Beneficiation Plant. The TCLD will consist of five 1,000 mm wide conventional belt conveyors. These conveyors will discharge onto the secondary crushed ore stockpile which will provide a buffer between the crushing operations and the beneficiation plant. Metal detection and magnets will be installed on the TCLD system to protect the Secondary screens and HPGRs. Electric hoists will be installed to service the various drive assemblies of the conveyors. Closed circuit television (CCTV) cameras will be installed to provide remote monitoring of transfer points.

Regularization (Plant Feed) Stockpile The secondary crushed ore stockpile will be conical, with a useful volume of 55,000 t (27,500 m³), to provide a live capacity of about 35 hours of operation of the HPGR circuit.

December 2019 120

The total stockpile capacity will be 220,000 t to enable a maximum of 140 hours of plant operation in the event of prolonged downtime within the primary and secondary crushing circuits. Four belt feeders will be used to feed the HPGR feed bins. It is expected that three of the feeders will be operating and one on standby. Metal detectors and extractors will protect the HPGR circuit from steel. CCTV cameras, remote from the control room will be used to monitor the stockpile reclaim system. Dust suppression equipment will be installed around the secondary crushed ore stockpile reclaim and conveying area.

High Pressure Grinding Rolls The ROM secondary crusher product (minus 55 mm) will be conveyed to the three HPGR feed bins. These feed bins will hold the new feed to the HPGRs as well as the circulating load of the HPGR crushing and secondary screening circuit. The feed bins each will have variable speed belt feeders which will transfer the ore to the HPGR in that circuit. The feed bin will have a live capacity of 6,600 t (3,300 m3) which is designed to provide a two hour operating surge capacity at the designed 100% circulating load. Each of the three HPGR roller presses will have 2.0 m diameter by 1.5 m wide rolls and be driven by two 1,800 kW motors. The product from the presses will be conveyed to a movable head conveyor which will discharge the pressed ore to one of four secondary screen feed bins. One of the biggest concerns regarding the operation of HPGR roller presses and their performance will be to keep the feed free of non-crushable materials that may damage their coating, which in addition to reducing their useful life and increasing operating cost, may compromise the efficiency of comminution. The feeders will be equipped with metal detectors which will activate flow diverter flaps. In this way, the feed containing metal will be diverted directly to the discharge conveyor bypassing the presses. CCTV cameras will monitor the feed bins and belt feeders as well as the belt conveyors linking the HPGRs and the secondary screens. The non-crushable material or metallic scrap will be collected together with the magnetite by the scrap extractor and discarded in a small pile that will be processed in due course. For environmental control, a baghouse dust collection system will be installed to handle feed and discharge areas of the presses. An 85/10 t overhead crane will be installed for the maintenance of roller presses. In addition, there will be secondary overhead cranage and hoists to service the conveyors and their drive systems.

December 2019 121

Secondary Screening, Ball Milling and Classification HPGR discharge will be stored in the four secondary screen feed bins. Each bin will have a live capacity of 4,700 t (2,350 m3) which will provide a 1.3 hour operational capacity for the secondary screening section. Belt feeders will transport the roller press product to the four secondary screens. The screens will be 12’ x 24’ double deck banana screens. A fifth screen will be available for replacement of any of the four operating screens when the need arises. The screen deck apertures will be 15 x 30 mm and 8 x 30 mm and the screens will be operated wet with spray wash water. The screen oversize (+8 mm) will be conveyed from the screens and returned to the HPGR feed bins. The design circulating load is 132%. The secondary screen undersize will flow into two pulp pumpboxes where it will combine with the overflow discharge from two 7.9 m (26 ft) dia x 12.2 m (44 ft) effective grinding length ball mills, each equipped with a 17,000 kW gearless drive. The grinding circuit classification will be carried out in two batteries of ten 26” hydrocyclones (seven operational and three reserves per battery). A design of 300% circulating load has been incorporated. The cyclone underflow will feed the two ball mills respectively. The grinding product will be returned to the pulp boxes from where it will be pumped back to the cyclone classification, closing the grinding circuit. A small fraction of this flow will be removed for process sampling. The overflow of the classification cyclone (80% passing 106 microns) will proceed to the flotation stage. A cyclone overflow sampling system with tubular samplers is planned. The sample will be sent to a particle size analyzer. In the grinding area, there will be numerous cranes and hoists installed to facilitate maintenance of the following:

• conveyors

• screens and belt feeders

• ball mills

• hydrocyclones

• hydraulic units

• cyclone feed pumps;

• conveyor vulcanizing press

The area will also have numerous CCTV monitoring points. The grinding area will have an inclined floor that will direct the overflow of boxes and possible leaks to the well located in the central grinding region. Sump pumps will return spillage to the grinding classification circuit. Any overflow from the milling area will be directed by gravity through a channel to the final flotation tailings line, with discharge at the tailings dam.

December 2019 122

The grinding media (balls with a maximum diameter of 63 mm) for the ball mills will be received in bulk in big bags, in trucks and stored in the concrete silo. The silo will be equipped with a rotary ball feeder whose function will be to remove the balls that will be transferred through the sidewall type high angle belt conveyor to feed the ball mills. The control of the dosage of the balls will be done through the weigh scale.

Rougher and Cleaner Flotation and Regrinding

Flotation Circuit The flotation circuit adopted will be very similar to that which is installed in the Salobo I & II lines, whose main characteristic is the joint flotation of coarse (solids with P80 of 106 microns) and fines (solids with P80 between 20-25 microns product of regrinding). Figure 48 is a schematic drawing of the Salobo III flotation circuit.

Figure 48 - Salobo III Flotation Circuit

Flotation Description The grinding product, after sampling will pass through two boxes with a trash screen. The output piping of each box will be unified and sampled through a tubular gravity sampler that will generate a sample for the online X-ray analyzer. It then will go to the sampling system, composed of a primary linear sampler, a secondary vezin and tertiary vezin, to remove the metallurgical samples, and finally, it will be conducted by gravity to a pulp distributor, where the flow will be divided for the two flotation lines.

FLOTAÇÃORG1

FLOTAÇÃOCL1

FLOTAÇÃOCL2

FLOTAÇÃOCL3

ESPESSAMENTO FILTRAGEM CLARIFICAÇÃO

FLOTAÇÃORG2

CLASSIFICAÇÃO/ REMOAGEM

FLOTAÇÃOSC-CL

FLOTAÇÃOSFR

ALIMENTAÇÃO

REJEITO FINAL

CONCENTRADO FINAL

(R)

(C)

(C)

(C)

(C)

(C) (C)

(C)

(R)

(R)

(R)

(R) (R)

(R)

December 2019 123

The Rougher I flotation will be developed in mechanical cells, of the tank type. For this stage of the circuit, two cells of 200 m³ are being installed each, one for each line, totaling a volume of 400 m³. The Rougher I concentrate will flow by gravity to the pulp boxes and will be sampled and analyzed by X-ray analyzers online. From the boxes, the rougher 1 concentrate will be transferred, to feed the Cleaner I stage. Alternatively, this concentrate could be fed to the Cleaner II feed or combined with Rougher II concentrate. The Rougher I tailings will flow by gravity to the Rougher II flotation, which will consist of two banks with six mechanical cells of 200 m³ each, totaling a volume of 2,400 m³. The Rougher II tailings will gravity feed the Staged Flotation Reactor (SFR) flotation, which will consist of two SFR cells. The SFR tailings will form part of the final tailings of the process. The pulp will be sampled for analysis on the online X-ray analyzer (one sample per bank) and, after joining with the tailings Scavenger of the Cleaner, will pass through the sampling system where a metallurgical sample will be collected by the primary linear sampler and secondary vezin and tertiary vezin units and will also pass through the trough sampler for analysis on the online X-ray analyzer. The final tailing will, by gravity, go to the Tailings Dam. The Rougher II concentrate together with the SFR concentrate, after sampling for online X-ray analysis will flow by gravity in parallel with the Cleaner Scavenger concentrate and then in a single pipe per line for the pulp pump boxes feeding the regrind circuit. The Cleaner I flotation will be done in flotation columns, with eight circular columns of 6.0 m in diameter and 14.0 m in height arranged in two rows of four, each group being fed by a pulp distributor with four outlets. The Cleaner I concentrate will be sampled and advanced to Cleaner II flotation. The Cleaner I tailings will feed Scavenger flotation (Cleaner Scavenger). The Cleaner Scavenger tailings will be sampled for online X-ray analysis. The Scavenger flotation of the Cleaner will be developed in two banks of five mechanical cells of 200 m³ each, totaling a volume of 2,000 m³. The Cleaner's Scavenger concentrate will be sampled for X-ray analysis and then sent by gravity for regrinding. The Cleaner Scavenger tailings will be part of the final tailings of the process. The pulp will be sampled for online X-ray analysis on (with one sample per bank) and after joining with Rougher tailings II, will constitute the final waste. The final tailing will be sampled by the linear primary sampler, secondary vezin tertiary vezin and, then, will be sent by gravity to the Tailings Dam through a concrete channel, which will also receive the drainage flow from the plant area including overflow from grinding, flotation, reagents and the recovered water tank. The Cleaner II flotation will also be in columns, with four columns with 4.3 m in diameter and 14.0 m high, arranged in two rows of two columns. The Cleaner II tailings will return to the Cleaner I flotation, joining the Rougher I concentrate. The tailings will be sampled by the pressurized pulp sampler for online X-ray analysis.

December 2019 124

The Cleaner II concentrate will proceed to Cleaner III flotation. Alternatively, this flow can be directed to the concentrate thickener feed box. The concentrate will be sampled for online X-ray analysis. The Cleaner III flotation will be carried out in two columns 4.3 m in diameter and 14.0 m in height. The Cleaner III tailings will return to the Cleaner II flotation and will be sampled by the pressurized pulp sampler for online X-ray analysis on the analyzer. The Cleaner III concentrate (final concentrate) will go to the concentrate thickening feed box by gravity. The final concentrate will be sampled by the chute sampler for online X-ray analysis on the analyzer and for the composition of metallurgical samples by the samplers.

Regrinding The classification of regrind circuit will be done in hydrocyclones of 15” in diameter. Two batteries of 14 hydrocyclones each, 12 operational and 2 reserves. The underflow of the cyclones will be directed, by gravity, to the vertical mills. Each receiving box of the underflow can feed the respective vertical mill of the line or the third central mill. Regrinding will be carried out by three vertical mills, with two being used from the existing plant’s lines I and II and one acquired (Metso) with power of 1,500 hp each. The discharge of the mills will be conducted by gravity to the pulp boxes joining the new cyclone feed and closing the overflow circuit. Operationally, two or three mills can be used. For this purpose, there is provision for maneuvering of dart-type valves in the UF channel of the classification cyclones and control valve to control the flow of feed to the third mill (central position). The overflow of the cyclones pulp with fine grained solids with P80 between 22 and 25 microns will be transferred to the Cleaner 1 flotation feed. For viewing and monitoring the products of the classification cyclone - overflow and underflow - Closed Circuit Television Systems - CCTV will be installed. For the ball loading routines of the mills, as well as for the maintenance of the equipment, a 10 t capacity crane will be provided. The regrinding area will be contained by small walls to direct the overflow of boxes and possible leaks to the inclined floor. The eventual overflow of the sump will be directed to the tailings channel at the end of the flotation.

Concentrate Thickening The function of concentrate thickening will be to dewater the copper concentrate, to improve filtration efficiency. The Cleaner III concentrate (final concentrate) will flow by gravity to the feed box of the 15 m diameter concentrate thickener.

December 2019 125

In addition to the Cleaner III concentrate, the thickener feed box receives flocculant through the reagent pumps. As an alternative to the flotation circuit, the box may also receive part of the Cleaner II concentrate, through the Cleaner III flotation distributor outlet, controlled by a pinch valve, directing this concentrate directly into the Cleaner III concentrate chute. This option aims to allow adjustment of the copper content of the final concentrate, when the ore has a higher Cu content. The concentrate will be sampled through the primary linear sample and by the secondary vezin. The sampling system will generate representative samples for the evaluation of the metallurgical and mass balance of the final concentrate. The overflow of the concentrate thickener will continue to overflow into the water tank. The recovered water will be used for regrinding. The underflow from the thickener, pulp with approximately 60% solids by weight, will be pumped to the pipeline or recirculated the pulp in the thickener, if the density is below the minimum required in filtration.

Concentrate Filtration and Storage The concentrate thickened by the existing plants (Salobo I and II) is transferred to the concrete pulp box. The existing box will also receive Salobo III concentrate. From the box, the thickened concentrate is transferred by gravity to the two existing filtration tanks. The tank has a useable volume of 1,200 m³ and will have the capacity for 36 hours of filtering operation. Its function will be to complement the regulating stock before filtering, guaranteeing the filters' operability, since they operate discontinuously, in cycles. The agitator will keep the pulp in suspension. From the existing surge tank, it is transferred to the filter presses. The filtrate generated in this pressing is collected and flows to the clarifier. After the filtering step, the cake is discharged directly onto the concentrate stockpile. The filtered cake, with an average humidity between 8.5 and 9.0%, is stored in a conical pile of about 1,100 t (related to the installation of the 5th filter). Considering the existing area, the capacity will reach a total of 3,750 t (relative to the filter discharge area). The filter cake is stored in a covered shed, under the filters. After unloading the cake, the screens and ducts are washed, and the washing effluent is incorporated into the filtrate and carried to the clarifier where the solids are separated from the water. The dense solids (clarifier underflow) are reincorporated into the filtration feed. The concentrate is reclaimed from the storage pile, by means of loaders, for trucks with a 35 t bucket. The trucks transport the concentrate from the processing facilities to the railroad concentrate loading facility in Parauapebas, about 125 km from the plant. The filtration building has an existing 7.5 t overhead crane for the maintenance of the filters, whose bearing path will be extended to meet the new building span.

December 2019 126

To visualize the operation of the filters, four cameras of the closed circuit television system - CCTV are installed, one for each filter.

Reagent Preparation

Collector 1 - Amil Xanthate (A350) This reagent will be received in solid form in a 700 kg big bag, and will be dosed in 8% aqueous solution (w / w), which will be prepared in batches. The amyl-xanthate, will be stored in closed sheds, for use for a period of 82 days (5 days of operation - stored in the shed of the processing plateau and 77 days of operation - stored in the warehouse shed). Each batch will correspond to approximately 45 h of operation. To prepare the solution, the big bags will be placed in the tank, with a volume of 60 m3, using the overhead crane with a capacity of 2.5 t. The reagent should be placed with the tank partially filled with water and with the agitator operating. After unloading the bags, the volume of water will be supplemented until reaching the desired dilution of 8% w / w. The solution will be transferred to the dosing tank, with a volume of 3.0 m3, by the pumps and from this tank to the batteries ball grinding cyclone, optionally for Rougher II flotation cells, to the Scavenger cells of the Cleaner and to the Scavenger distributor of the Cleaner, and optionally for distributors of Cleaner I cells. For environmental control, a containment basin will be installed in order to contain any leaks, where the solution contained in the basin may be returned to the preparation tank through the portable pneumatic recovery pump. Pipes and quick coupling connections will be provided to allow the reagent to be recovered into the tank. An overhead crane is planned to serve the entire reagent area.

Collector 2 - Hostaflot 7800 Collector 2 will be received in tank trucks and transferred by pumps to the storage tank with a volume of 150 m³. The truck unloading system will be unique and will also serve the reagents Frother 1 and 2 and Collector 3. From the tank the solution will be transferred by the pumps to the dosing tank, with a volume of 0.2 m3, and this tank for the ball milling cyclones, optionally for Rougher 2 flotation cells for the Scavenger Cleaner cells, for the Scavenger Cleaner distributor and for the Cleaner II pulp distributor. For environmental control, containment basins will be installed in order to contain any leaks.

December 2019 127

Frother 1 - Flotanol M28 Flotanol will be received in liquid form, in tank trucks, and will be used as received (without dilution). From the trucks, the reagent will be transferred and stored. The tank will have a volume of 180 m³, with a range of 110 days. The flotanol will be dosed without dilution and will be transferred from the tank to the dosing tank with a volume of 0.2 m3, and from this tank to the boxes, to the Rougher II flotation cells and optionally for the Rougher II cell for the Cleaner Scavenger distributor and optionally for Scavenger from Cleaner. For environmental control, containment basins will be installed in order to contain any leaks.

Frother 2 - Polypropylene Glycol Methyleter (Flotanol D 25BR) Flotanol D 25BR will be received in liquid form, in a tank truck, and will be used as received (without dilution). The reagent will be transferred to the storage tanks by the pumps. The tank will have a volume of 180 m3, with a capacity of 88 days. From the storage tank the reagent will be transferred to the distribution tank, with a volume of 0.2 m3, by the pumps and this tank optionally for the Cleaner Scavenger distributor and optionally for Cleaner I distributors, for Rougher II and optionally to Cleaner II. For environmental control, containment basins will be installed in order to contain any leaks.

Collector 3 (Hostaflot 7800, or similar) Collector 3 will be received in the form of a 100% concentration solution, and will be dosed in the same receiving concentration. Collector 3 will be received in tank trucks, and transferred to the 240 m³ storage tank. From the storage tank the solution will be transferred to the dosing tank, with a volume of 6.2 m3, and from this tank for the ball milling cyclone batteries, optionally for the Cleaner I, to the pulp distributor of the cleaner scavenger cells, through the pump, to cleaner scavenger cells, optionally for Rougher II flotation cells and optionally to the Cleaner II. For environmental control, containment basins will be installed in order to contain any leaks.

December 2019 128

Flocculant The flocculant will be received in solid form, in 25 kg bags, and will be stored in a closed shed, for use for a period of 150 days (stock of 39 days in the shed of the plateau and 111 days in the warehouse). The preparation of the flocculant will be done in two stages. In the 1st stage, a 0.3% stock solution will be prepared in batches, which will then be diluted in line to the 0.03% concentration. The stock solution will be prepared in the tank, which is equipped with the agitator. The powdered flocculant will be transferred manually to the receiving silo. From this silo, the flocculant will be dosed, through a system of a rotary valve and ejector in the tank, which it should be partially filled with water and with the agitator operating. After the transfer of the flocculant, the volume of water in the tank will be complemented, until reaching the desired dilution of 0.3% w / w. From the tank, with a volume of 20 m3, the flocculant stock solution will be transferred to the 3.0 m3 distribution tank, through pumps. These pumps will be of the positive displacement type (screw), due to the high viscosity of this reagent. From the distribution tank, the flocculant will be dosed through the metering pumps in the concentrate thickener supply box. In the discharge of these pumps, the 0.3% flocculant solution will be diluted to 0.03%, for better performance of this reagent, through a line mixer. For environmental control, containment basins will be installed in order to contain any leaks, where the solution contained in the basins may be directed to the tailings dam.

Lime The lime will be received in big bags of 1,250 kg, and stored in a closed shed, for use for a period of 5 days in the reagent shed. Alternatively, the lime can be received by a bulk truck, with compressors for pneumatic transportation. The bags will feed the 60 m³ silo, manually, through the overhead cranes. For the environmental control of this operation, the dust collection system is provided, with a bag filter and an exhaust fan. The preparation of lime milk (10% suspension) will be done in batches. The lime will be dosed by the dosing scale in the tank, with a volume of 8 m3, equipped with an agitator. Dilution water will be added to this tank, with its flow controlled by the scale, to ensure the desired concentration is obtained. From this tank the lime milk will be transferred by the pumps to the distribution tank equipped with an agitator, with volume 55 m³. From the tank, the lime milk will be distributed by the pumps to the consumption points of the plant,

December 2019 129

which will be: optionally for underflow of the milling cyclones, for the Rougher II flotation cell and also in the distributor supplying the Cleaner Scavenger flotation stage. For environmental control, a basin and low walls will be installed in order to contain any leaks. A well with a vertical pump will also be provided. The solution can be recovered into the lime storage tank. The eventual overflow will be directed to the final waste channel. Two overhead cranes with a capacity of 2.0 t will be provided.

17.3. Energy, Water, and Process Materials Requirements The plant is provided with electricity from the plant substation. Step-down transformers provide the various voltages used by the equipment. The bulk of the process water needs are covered by the recirculation from the TSF. The consumption of fresh water is limited to systems requiring such a quality. Water is provided by vertical pumps installed in the Mirim and Salobo Creeks. Reagent dosages, as budgeted for the 2015–2019 period, are 71 g/t for PAX, 60 g/t for dithiophosphate, 70 g/t for propylene glycol and 90 g/t for MIBC. NaHS and lime consumptions are at 120 g/t and 600 g/t, respectively. The other major consumables are the grinding balls, with the ball mills calling for 600 g/t of the 76 mm balls, and the regrinding stage 50 g/t of 20 mm balls.

17.4. Plant Equipment and Design Considerations The major process equipment installed in the Salobo I and Salobo II sections of the plant are shown in Table 41 as well as the planned equipment for Salobo III.

December 2019 130

Table 41 - Major Process Equipment

Salobo I Salobo II Salobo III

1 Gyratory crusher – 60 in. x 89 in. 1 Gyratory crusher – 60 in. x 89 in.

1 Gyratory crusher – 60 in. x 89 in.

2 Cone crushers 1 Cone crusher 2 Cone crushers

2 Vibrating screens 12 x 24 ft 1 Vibrating screen 12 x 24 ft 2 Vibrating screens 12 x 24 ft

1 Overland pipe conveyor (78.7 in.); 6,000 hp; 1,700 m length; 4,600 t/h capacity

Equipment shared with Salobo I 5 conventional conveyors, 2286 t/h capacity

2 High pressure grinding rolls – Ø 2.0 x 1.5 m 2 High pressure grinding rolls – Ø 2.0 x 1.5 m

3 High pressure grinding rolls – Ø 2.0 x 1.5 m

2 Ball mills – Ø 26 ft x 40 ft 2 Ball mills – Ø 26 ft x 40 ft 2 Ball mills – Ø 26 ft x 40 ft

24 Flotation tank cells – 200 m3 24 Flotation tank cells – 200 m3 24 Flotation tank cells – 200 m3

14 Flotation columns – 14 m 14 Flotation columns – 14 m 14 Flotation columns – 14 m

2 Flotation SFR – 60 m³ 2 Flotation SFR – 60 m³ 2 Flotation SFR – 60 m³

3 Vertimills – 1,500 hp 3 Vertimills – 1,500 hp 3 Vertimills – 1,500 hp

1 Concentrate thickener – Ø 15 m 1 Concentrate thickener – Ø 15 m

1 Concentrate thickener – Ø 15 m

2 Pressure filters 2,000 x 2,000 / 36 chambers 2 Pressure filters 1,500 x 1,500 / 50 chambers

1 Pressure filters 1,500 x 1,500 / 50 chambers

Note: * One cone crusher and one vibrating screen are common for both Salobo I and II plants, as stand-by equipment;

Ø = effective grinding length.

Salobo I relied on a series of conveyors that were oversized for the design capacity of the 12 Mtpa plant, but were planned to cover the eventual addition of Salobo II and the total 24 Mtpa expanded capacity. The resulting processing facilities therefore have a series of critical items that are not duplicated, the failure of which can curtail operation from both Salobo I and II, if the failure is located after the plant stockpile where no more significant surge capacity exists. Figure 49 shows the process lines and highlights in grey the elements found as single-line items in the flowsheet. It shows that 12 conveyors, including two with shuttle heads, form a critical path after the last available stockpile. This presents a risk for interruptions in plant operation however the improvements in plant availability and utilization demonstrate that these items have been addressed.

December 2019 131

Figure 49 - Process Plant Single Line Equipment


17.5. Process Plant Performance Projections The copper and gold recoveries applied to 2020 to 2023 production plans were defined based on historical plant performance data including main factors that can affect the recovery in the mineral processing plant. The plant performance projections are shown in Table 42. From 2023 onward, the processing recoveries used in LoMP are shown in Table 43. Both of these tables have been detailed in Section 13.5.3.

Table 42 - Processing Recovery Assumptions (2020-2023)

Metal 2020 Total Recovery to All Concentrates (%)

Copper -3.154*(1/A) + 90.6108*(-0.0745*B + 1.2321) - (3.96* C)

Gold -59.421*(D^3) + 152.37*(D^2)-103.9*D + 93.388*(-0.0745*B + 1.2321)- (2.21* C)

Metal 2021-2023 Total Recovery to All Concentrates (%)

Copper -3.154*(1/A) + 90.6108*(-0.0745*B + 1.2321) - (3.96* C)

Gold -59.421*(D^3) + 152.37*(D^2)-103.9*D + 91.815*(-0.0745*B + 1.2321)- (2.21* C)

Notes: Factors A=Feed Cu grade (%), B=Wet Plant feed rate (kmt/hour), on the basis of two lines of operation C=Percentage in mass of transition ore (%) and D=Feed Au grade(g/t)

December 2019 132

Table 43 - Processing Recovery Assumptions (2024-LoMP)

Metal Total Recovery to All Concentrates (%)

Copper (-2.5362*1/%Cu in feed ) + 90.674 Gold (1.0173*RecCu) - 20.357

Table 44 - Processing Plant Performance Forecasts

2020 2021 2022 2023 2024

ROP (wmt) 24,338,943 24,333,511 33,807,407 36,604,813 36,734,694

Cu Feeding Grade (%) 0.97 0.96 0.85 0.83 0.69

ROP (dmt) 23,852,164 23,846,841 33,131,258 35,872,717 36,000,000

Au Feeding Grade (g/t) 0.63 0.58 0.45 0.45 0.37

Concentrate Production (dmt) 528,262 526,238 626,484 678,555 565,593

Concentrate Production (wmt) 590,237 587,975 699,982 758,161 631,947

Concentrate Moisture (%) 10.5 10.5 10.5 10.5 10.5

Concentrate Cu Grade (%) 38 38 38 38 38

Contained Cu in Concentrate (t) 200,739 199,970 238,064 257,851 214,925

Contained Au in Concentrate (kg) 10,941 9,835 10,095 11,405 9,079

Cu Recovery (%) 87.1 87 85 86.4 87

Au Recovery (%) 73.4 71 67.7 70.1 68.1

Concentrate Au Grade (g/t) 20.7 18.7 16.1 16.8 16.1

The application of the equations in Table 11 for 2022 are further modified on an overall basis due to the planned ramp-up of Salobo III. Plant throughput and metallurgical performance of both copper and gold have been modeled for the first year of Salobo III operation based on the most recent commissioning and ramp-up of Salobo II in H2 of 2014 and H1 of 2015. Throughput ramp-up of Salobo III has been modeled to continue into 2023. Table 45 presents the actual recent historical operating recoveries and a comparison to the forecasted recovery assumptions.

Table 45 - Actual and Forecasted Processing Recoveries

Actuals Vale 5Y Forecast

2017 2018 2019 2020 2021 2022 2023 2024

Mill Feed kWMT 24,197 24,224 23,033 24,339 24,334 33,807 36,605 36,735

Mill Feed kDMT 23,650 23,657 22,486 23,852 23,847 33,131 35,873 36,000

Mill Feed DMT/hr 3,078 3,038 3,002 3,087 3,107 4,562 4,672 4,687 Mass

Recovery % 2.1 2.2 2.3 2.2 2.2 1.9 1.9 1.6

Cu Recovery % 88.3 86.1 86.9 87.1 87 85 86.4 87

Au Recovery % 69 71.7 74.4 73.4 71 67.7 70.1 68.1 Cu

Concentrate kDMT 498 510 510 528 526 626 679 566

Cu Produced kt 193 193 190 201 200 238 258 215

Au Produced kozs 346 361 368 352 316 325 367 292

December 2019 133

17.6. Comments on Section 17 In the QPs’ opinion, the Salobo concentrator is currently operating at throughput and performance levels upon which it was designed. Plant availability, utilization and throughput have been at design levels over the past 12 months. Metallurgical performance for copper and gold have improved and are currently near or at design levels on a continuous basis and performance has exceeded design at various periods. The operating team have made continual modifications and improvement to the plant which has resulted in the improvements to availability and performance with the installation of the SFRs as an example of this. Efforts are ongoing to utilize a geometallurgical program to further optimize plant recoveries and recent improvements in gold recovery have resulted. The QPs agree with the use of dedicated metallurgical samplers in the reject and concentrate lines of the plant. This initiative is currently being implemented at both Salobo I & II and Salobo III and should result in an improvement in the precision of the metallurgical balance.

18. Project Infrastructure

18.1. Roads and Logistics The area is well-served by railroads and highways that connect the towns and cities. Regularly scheduled air service is available in Marabá approximately 225 km from Salobo by highway. Most flights connect to the capital, Brasilia. Vale has a contract with a transportation company to transport all proper employees and contractors in Carajás and Parauapebas to the Salobo Operations site. The traffic routes are elaborate and supervised by Vale. Employees also use a fleet of company owned vehicles for transportation. For safety, all the DIMB vehicles have a system that registers speed, abrupt stops, and rpm. The system signals the driver when the vehicle is travelling above the allowed speed. Copper concentrate is shipped by truck (40t) from the Salobo Operations to a rail facility near the city of Parauapebas, 85 km from the mine. From there it is transported by rail to the Ponta da Madeira Marine Terminal at São Luís, a distance of approximately 870 km, for shipment.

December 2019 134

Figure 50 - Concentrate Load Out, Parauapebas

Source: Google Earth

18.2. Stockpiles and Waste Rock Storage Facilities Low-grade ore and waste rock from the mine are stored in three locations along the perimeter of the pit (Figure 51). The main WRF is to the west of the pit, and contains both oxidized and fresh rock. Additional information is included in Section 20.

December 2019 135

Figure 51 - Location Plan, Waste Rock Storage Facility and Low-Grade Stockpiles

18.3. Site Infrastructure

Surface facilities include:

• Central Administrative Facilities, which includes administrative offices, restaurant, change

rooms, training centre and a medical clinic

• Central Maintenance Facilities, which includes a mine heavy equipment workshop

including tyre changing facility, a light vehicle maintenance shop, a plant maintenance

shop for component overhaul and repair, a warehouse and maintenance offices

• Mine facilities, which includes mine operations change rooms and mine operations offices

• Mine heavy equipment fuelling facilities, which are located next to the primary crushers

• Main substation

• Small vehicle fueling station

• Recycle centre

• Security/ access control gate

December 2019 136

18.4. Tailings Storage Facility The Salobo tailings storage facility (TSF), comprising an earth dam and concrete-lined spillway, was designed for Vale by Brazilian engineering company BVP Engineering to withstand a 1 in 10,000 year event. The current design was completed in July, 2010. The TSF, when completed to an elevation of 285 m, will have sufficient capacity to store the entire forecasted tails from the processing of the Mineral Reserve planned over the life of the mine. Potentially the TSF site could also store the forecasted tails from the material presently identified as Mineral Resources. During 2019 the dam was raised from 245 m elevation to an intermediate design height of 255 m elevation. The Tailings Storage Facility is shown in Figure 52 and its storage capacity is shown in Figure 53.

Figure 52 - Salobo Tailings Facility with 2019 Lift Completed

December 2019 137

Figure 53 - Tailings Storage Capacity and Area versus Impoundment Height

18.5. Water Supply Process make-up water comprises runoff and direct precipitation within the tailings storage basin. This raw water is pumped to the plant together with return water from tailings deposited in the storage facility. If the plant requires additional makeup water, this can be extracted from Igarapé Mamão (Pawpaw creek) via a floating intake within the project site, using vertical pumps.

18.6. Power and Electrical In accordance with legislation governing the Brazilian electrical power sector, the Salobo Operations are supplied by the Eletronorte division of Eletrobras, responsible for the northern region of Brazil, operating and maintaining the system on behalf of the National Operator of the Electrical System (NOS). Electrical energy is supplied from Tucuruí, an 8,370 MW hydroelectric generating station on the Tocantins River, 200 km north of Marabá, and 250 km due north of Parauapebas. The 150MW of power required by the Salobo Operations is transmitted 87 km by an overhead 230 kV transmission line.

18.7. Communications Telephone communications are available over landlines, and via a cellular network. Internet communications are also available at the mine site.

December 2019 138

18.8. Housing Local housing is available for employees within the Carajás urban centre and Parauapebas. There are adequate schools, medical services and businesses to support the work force. The mine sites have medical facilities to handle certain emergencies. In addition, medical facilities are available in to support the mine’s additional needs. Vale has invested significantly in the town’s infrastructure, building a 130 km paved road to Parauapebas and a 20 km sewage system, together with a school, hospital, and daycare centre.

18.9. Comments on Section 18 The Salobo mine has been in operation since 2012, and at has been producing at a 24 Mtpa capacity since 2014. The infrastructure is well established and adequate for this production level and the expanded 36 Mtpa level planned for in the LoMP.

19. Market Studies and Contracts Vale has agreements at typical copper concentrate industry benchmark terms for metal payables, treatment charges and refining charges for concentrates produced. Treatment costs and refining costs vary depending on the concentrate type and the destination smelter. For all of Vale’s sales contracts, the risk of the concentrates transfers either at the load port or discharge port according the standard International Commercial Terms (Incoterms); whereas the title to the concentrates transfers either at the load port or discharge port according the standard Incoterms or upon payment. The terms contained within the sales contracts are typical and consistent with standard industry practice and are similar to contracts for the supply of copper concentrate throughout the world. Depending on the specific contract, the terms for the copper concentrate sale are either annually negotiated, benchmark-based treatment and refining charges, or in the case of spot agreements are based on fixed treatment and refining charges based on market terms negotiated at the time of sale. The differences between the individual contracts are generally in relative quantity of concentrates that are covered under annually-negotiated treatment and refining charges. The typical grade of copper, gold and silver in the final product is approximately 38%, 20 g/t and 75 g/t, respectively. The copper concentrate is used in copper smelting and refining operations to produce copper cathode and precious metals. Markets for copper concentrates is well developed with a large number of custom smelters located around the world who use the copper concentrate as feed. Higher levels of fluorine, higher copper grade and other specificities limit some of the processing options for the Salobo concentrate. Customers for Salobo concentrate have been well established.

December 2019 139

The Wheaton streaming agreement is discussed in Section 4.4. Metal price and exchange rate assumptions are shown in Table 46.

Table 46 - Metal Sale Price and Exchange Rate Assumptions

Item Units 2020 2021 2022 2023 2024 Long Term

Copper Price US$/tonne Cu 6,125 6,200 6,250 6,600 7,100 7,000

Gold Price US$/oz. Au 1,325 1,350 1,350 1,370 1,360 1,290

Exchange Rate Real$/US$ 3.8 3.8 3.8 3.8 3.8 3.8

19.1. Comments on Section 19 It is the QP’s understanding that the Salobo concentrates have been successfully marketed to custom smelters located around the world. A definition drilling campaign was initiated to increase the confidence of the Cu and Au grade variability and grade distribution of deleterious elements (uranium, fluorine, and chlorine) and their impact on the copper concentrate quality will also be used for silver evaluation and its applicability in the long-term model and mine planning. This drilling will assist in improving the confidence of the mineral reserves and ability to deliver the LoMP grades.

20. Environmental Studies, Permitting and Social or Community Impact

Environmental and social baseline study areas were defined to characterize the current conditions in the areas potentially affected by mine components or activities. The project is located in the Carajás mountain range in the eastern Amazon humid tropical rainforest. Temperatures range from 20.8°C to 37.8°C with an average relative humidity of 80.5%. Mean annual rainfall is 1,920 mm and evaporation is 1,500 mm. Winds are predominantly from the north and west. The project lies in part of the Salobo Creek and the Cinzento River basins which are tributary to the Itacaiúnas River. The long-term average unit runoff for the project site is 2.02 m3/s. The Tapirapé–Aquiri National Forest has a registered area of 190,000 ha. The Tapirapé Biological Reserve, which covers an area of 103,000 ha, borders the National Forest (and mine area) to the north. The mine site is within the Tapirapé–Aquiri National Forest and the access road crosses the Carajás National Forest and lies adjacent to the Igarapé Gelado Environmental Protection Area. Figure 54 shows the location of the mine in relation to the forest areas.

December 2019 140

Figure 54 - Mine Location in Relation to Forest Reserves

Note: Figure courtesy Vale, 2015. Floresta Nacional = national forest, Reserva Biologica = biological reserves’ Rio = river; APA = protected area.

As a condition of the mine installation license, an agreement was signed between the Chico Mendes Biodiversity Conservation Institute and the Salobo Operations to provide payment and support towards management of the Tapirapé–Aquiri National Forest (ICMBio, 2007). The protected areas have distinct management categories that were established by Decree N° 97,720 dated 5 May 1989. Within these areas, a regular polygon outlining the mining zone Special Use Area was defined by the National Department of Mineral Production of Brazil. Within the Special Use Area, Vale controls access to the area and the mine site, and access to the Tapirapé–Aquiri National Forest along the eastern boundary of the Special Use Area with the forest. To the northwest of the Special Use Area is the Lindoeste settlement, developed on land in the São Felix do Xingu region, which currently covers about 120 ha; the mine site has no influence over forest access by this community. Salobo Operations are also committed to offsetting the effects of its operation by planting seedlings of native species in proportion to the vegetation suppression carried out.

December 2019 141

20.1. Environmental Management The Salobo Operations have an Environmental Control Plan (Brandt, 2003) that includes the following components:

• Project Description

• Environmental Management System

• Vegetation Clearing and Stripping

• Erosion Control

• Water and Effluent Management

• Waste Management

• Atmospheric Emissions

• Noise and Vibration

• Environmental Emergencies

• Disease Control

• Archaeology Protection and Salvage

• Rehabilitation Plan

• Environmental Compensation and Social Inclusion

• Environmental Education

• Environmental Monitoring

• Closure Plan These social and environmental management plans detail best practices and Brazilian legislation to prevent and mitigate potential impacts and manage compliance specifically for the Salobo Operations.

20.2. Permitting Brazil is a federal republic, and its legal system is based on Civil Law tradition, characterized by codification of legal requirements. The Federal Constitution (October, 1988) is the basis of the legal system. Key applicable legislation for construction, operation and closure of the Project includes the following:

• Mining Code (Decree-Law No. 227, 28 February, 1967) and its Regulations (Decree No. 62934, 2 July, 1968)

• Forest Code (Law 12.651, May 25, 2012)

• National Environmental Policy Law (Law No. 6938, 31 August, 1981)

• CONAMA (National Environment Council) Resolutions Nos. 1/86, 23/86, 9/90, 10/90 and 237/97; and

• National Water Resources Policy Law (Law No. 9433, 8 January, 1997).

• National Law on National Solid Waste Policy (Law No. 12,305, August 2, 2010).

December 2019 142

The Preliminary License No. 33/94 was issued for the Salobo project in September 1994 following preparation and submission of the Environmental Impact Analysis/Report on Environmental Impact (EIA/RIMA) in 1992. The implementation of Salobo started in 2010 and received its first Operating License No. 1096/2012 on November 5, 2012. The current license refers to the research, mining and mineral processing of 24 Mtpa as well as all administrative and support facilities, including workshops, the central material disposal area and warehouse, dining hall, transportation, storage and shipment of copper concentrate, which is valid until October 19, 2024. The Installation License for the expansion of the feed stockpile from Salobo processing plant to 36 Mtpa (No. 1249/2018) valid until November 22, 2022. Regarding the removal of vegetation, Salobo has five valid licenses: Nº 1181/2016 (254,448 ha) valid until November 23, 2021; No. 1188/2017 (574.68 ha) valid until February 21, 2021; No. 10539201917636/2019 (823.71 ha) valid until December 3, 2021; and those of Nº 1001/2015 (408.47 ha) and Nº 1104/2016 (48.15 ha) that are under renewal with protocol requests on time (Protocol - 02001.014 542 / 2016-51 SEI nº 5276794 and Protocol - SEI n ° 4029492). Salobo also has other Operation Licenses related to service stations for light and heavy vehicles (nº 1035/2011 and 1081/2011, respectively) and to the Parauapebas copper storage railway station (nº 12083/2023) valid until February 20, 2023. The Operation License for the heavy vehicle filling station is valid until November 5, 2021. The light vehicle filling stations (valid until June 20, 2016) and has been renewed within the validity period (protocol nº 02001.002794 / 2016-37 respectively). The Brazilian government does not have a deadline to renew the license, however, as the renewal was requested within the validity period, the current license is still considered valid. Salobo Operation also has three valid installation licenses: The Installation License No. 1046/2015 refers to the expansion of the feed stockpile at Salobo processing plant to 24 Mtpa, which semi-annually it is informed to Brazilian Government about the geotechnical stability and efficiency of the control systems. The Installation License No. 1157/2017 for the heightening of the Salobo Dam up to the level 255 m is valid until August 17, 2020. The Installation License No. 1209/2018 for implementation of pumping systems of fines containment dikes is valid until April 10, 2020 but has been renewed by protocol No. SEI 6856994. Regarding the Salobo III Project, three environmental licenses were issued: Installation License No. 1249/2018 for the expansion to 36 M tonnes per year, Vegetation Removal License No. 1339/2018 and Authorization to Capture, Collect and Transport Biological Material No. 1017/2018. These licenses concern the installation of a sulfide copper ore beneficiation plant and its associated infrastructure. There is also the surface water capture and discharge concession (No. 1896/2017) granted in October 9, 2017 and valid until October 9, 2027, and the underground water

December 2019 143

capture concession for explosive factoring (No. 2519/2016) granted on June 17, 2016 and valid until May 16, 2020 which was renewed by protocol No. 2019/47490. The Salobo Operations currently hold all required permits to operate. The mine has a robust control and monitoring system to ensure that permits remain current, and to ensure that the requirements of each permit are monitored to comply with the relevant regulatory conditions imposed.

20.3. Social and Community Impact The Salobo Operations area of influence is the southeast Paraense mesoregion, in the municipalities of Marabá and Parauapebas. These regions are considered to have moderate human development indices for the level of health, education and living conditions, based on data from 2000. The extractive industry accounts for 23.5% of the economic activity in the state of Pará, with 17.9% other industrial activities, 52.0% services and 6.6% farming and ranching based on 2010 data (IBGE, 2013). The Project is not located on indigenous lands. The nearest indigenous lands include the river Tapirapé Tuere, Trincheira Bacaja and Xicrin do Cateté, all located 25 km or more from the Project. The Xikrin indigenous peoples traditionally use the Project area for food collection. In 2001, a forest management program was implemented between the indigenous communities and government associations to sustainably harvest the forest in the Project area in a manner that benefitted the indigenous community in capacity building and financial resources. Vale currently maintains a Communication Plan that commits to continued communication with the local indigenous to maintain community health and safety, cultural preservation, transparency of activities and harmony between the workers and the indigenous community. There are a number of social management plans carried out by the Social Communications Department. The Environmental Compensation and Social Inclusion plan objectives are to support sustainable development by capitalizing on the positive effects of project development and minimizing the potential negative effects. In addition, this plan is supported by a Social Communications program that facilitates information exchange and works to improve relations between the Salobo Operations and the diverse social segments of the surrounding communities. An Environmental Education program was developed at the Sustainability Department. The program seeks to disseminate the principles of sustainability recognized as environmental, social and economic responsibility through educational activities aimed at Vale employees and contractors and the surrounding community. The program aims to strengthen and expand environmental education in the municipal education program and in the community.

December 2019 144

20.4. Closure Plan The mine Closure Plan assumes that there will be partial recovery of infrastructure for use by educational activities, research and tourism. The closure plan is included in the Environmental Control Plan and rehabilitation and re-vegetation work is ongoing during operations. The Closure Plan (SETE, 2015) outlines the steps to be taken for the progressive rehabilitation and ultimate closure of the open pit and concentrator facilities and the auxiliary components of the operation and all associated infrastructure and equipment. The Closure Plan will be reviewed in 2020. The overall objective is to return the Project area to a natural condition to support the local vegetation and wildlife biodiversity of the Tapirapé–Aquiri National Forest. There are no reclamation bonds required for the mine. Rehabilitation and re-vegetation work are ongoing during operations. Closure costs have been estimated by Vale at approximately US$110 million. Closure costs are to be reviewed annually and are included as indirect costs in each operational centre for budgeting, expenditure tracking and financial planning (Vale, 2019).

20.5. Comments on Section 20 It is the QPs understanding that the Salobo Operations have all required permits for operating and that appropriate environmental management systems are in place. Also, that an approved closure plan is in place designed to return the site to a natural condition to support the local vegetation and wildlife biodiversity of the Tapirapé–Aquiri National Forest.

21. Capital and Operating Costs

21.1. Capital Cost Estimates The historical and estimated sustaining capital estimates are presented in Table 47 and Table 48, respectively.

Table 47 - Historical Capital Expenditures Summary (US$ M)

Type 2012 2013 2014 2015 2016 2017 2018

Growth 700.3 446.1 368.8 52.9 0 0.1 2.9

Sustaining 67.8 58.8 54.7 150.3 150.3 77.05 123.1

In 2019 a total of US$113 M is expected to be invested in sustaining capital, this amount is composed mainly of dam works (US$17 M), Mine Equipment Spare Parts (US$12 M), acquisitions and replacement of mine equipment (US$9 M), construction of new administrative structures (US$9 M) and vegetation clearance (US$9 M).

December 2019 145

A total of US$756.3 M will be invested in sustaining capital over the next five years for mine and processing plant improvement and upgrades (equipment, materials, spare parts, etc.), health, safety, and environmental sustaining expenditures relating to dam works. In 2020, the sustaining capital is composed mainly of acquisitions and replacement of mine equipment (US$27 M), dam works (US$25 M), construction of new administrative structures (US$13 M), construction of roads (US$10 M), plant maintenance components (US$8 M) and Mine Equipment Spare Parts (US$7 M). The Salobo III Project construction began in 2019 with production ramp-up planned for 2022. The total capital expenditure is estimated to be US$1,022 M over four years. The Salobo III Project is expected to increase processing plant capacity at Salobo to 36 M tonnes per annum in 2022.

Table 48 - Capital Expenditures Summary (US$ M)

Item

2018 20191 2020 2021 2022 2023 2024

Actual Forecast

Growth – Salobo III 2.9 143.7 401.1 368.6 105.7

Sustaining 123.3 113 153.4 168.6 221.5 200.4 111.1

Mine 98.4 62.5 59.5 63.4 98 105.8 72.2

Processing Plant 19.9 20.2 21.6 40.5 24.4 17 11.2

Dikes and Dams 2.9 16.4 28.7 22.1 72.3 73.3 24.2

Logistics 0.5 1.3 6.7 7.4 4.3 2.3 1.4

Others 1.6 12.6 36.9 35.2 22.5 2 2.1

Total (US$ M) 126.2 256.7 554.5 537.2 327.2 200.4 111.1

Note1 - Full cash operating costs per tonne Cu – excludes only amortization and depreciation

21.2. Operation Cost Estimates Table 51 shows the historic operating costs from 2015 to 2018.

Table 49 - Historic Operating Costs (US$ M)

Item 2015 2016 2017 2018

Mine 305 311.7 346.2 334.3

Processing Plant 199.5 183.2 209.4 186.8

Other Costs 71.7 64.9 77.6 78.3

Total 576.1 559.8 633.3 599.4

Tonnes Cu 155,391 175,923 193,380 192,592

Cost per Tonne Cu 3,708 3,182 3,275 3,112

Although higher costs were incurred for maintenance materials, fuel and tires, the annual unit mining operating cost decreased in 2019 compared to 2018 due to changes in the BRL / US$ foreign exchange rate. Table 50 lists the forecasted LoMP unit costs. Total annual costs are divided by total mine movement to determine the operating unit costs as an output.

December 2019 146

Table 50 - Forecast Operating Cost

Cost (US$ M) 2018 2019

2020 2021 2022 2023 2024 Actual Forecast

Mine 334.3 341.9 340.1 346.1 348.2 350.7 343.5

Processing Plant 186.8 175.9 188.1 193.4 269.3 282 284.8

Other Costs1 78.3 74.8 85.2 83.7 91.3 97.5 89.6

Total 599.4 592.6 613.4 623.2 708.8 730.2 717.9

Tonnes Cu 192,592 194,364 200,739 199,970 238,064 257,851 214,925

Cost per Tonne Cu 3,112 3,049 3,056 3,116 2,977 2,832 3,340 Note1 - Excludes Ocean Freight

21.2.1. Mine Operating Costs The operating cost estimation is performed in conjunction with the mobile equipment fleet selection and mine planning. In addition to the equipment’s direct operating costs, the other key factors include labour, salaries, energy, and fuel costs. Table 51 lists the forecasted LoMP unit and operating costs. Total annual costs are divided by total mine movement tonnes to determine the operating unit costs as an output. The operating cost estimation is completed in conjunction with the equipment fleet selection process. In addition to the equipment direct operating costs, other key factors include labor, salaries, energy and fuel costs.

Table 51 - Mine Operating Costs

Item 2018 Actual 2019 Forecast 2020 2021 2022 2023 2024

Cash Cost (US$ M) 334.3 341.9 340.1 346.1 348.2 350.7 343.5

Mined Material (wet M t) 140.8 138.9 140 140 140 136 126

Unit Cost (US$/tonne) 2.4 2.5 2.4 2.5 2.5 2.6 2.7

21.2.2. Process Operating Costs Annual unit processing costs were reduced in the actual 2019 LoMP compared to the 2018 LoMP. Table 52 lists the estimated process plant operating costs.

Table 52 - Process Operating Costs

Item 2018

Actual 2019 Forecast 2020 2021 2022 2023 2024

Cash Cost (US$ M) 186.8 175.9 188.1 193.4 269.3 282 284.8

Processed Material (dry Mt) 23.7 23.1 23.9 23.8 33.1 35.9 36

Unit Cost (US$/t) 7.9 7.6 7.9 8.1 8.2 7.9 7.9

21.2.3. Other Operating Costs The Royalties (CFEM – Federal Royalty – Mineral Exploitation Levy) are calculated on the net revenue obtained at the time the mineral product is sold. For copper the CFEM rate is 2%. For calculation purposes, net revenue is considered as the sale value of the

December 2019 147

mineral product after deduction of the taxes levied on its commercialization, transport and insurance. Table 53 lists the estimated other costs.

Table 53 - Other Costs

Item 2018 Actual 2019 Forecast 2020 2021 2022 2023 2024

Cash Cost (US$ M) 78.3 74.8 85.2 83.7 91.3 97.5 89.6

21.3. Comments on Section 21 In the QPs opinion, projected operating and capital costs are reasonable as they are in line with historical actuals.

22. Economic Analysis With the Salobo III expansion, Wheaton is contractually obliged to make an expansion payment once the new processing line has achieved completion (Expansion Payment). The Expansion Payment is based on a matrix which varies depending on three variables: (1) the date on which completion of the Salobo III expansion is achieved; (2) the size of the Salobo III expansion; and (3) the copper cut-off grade for direct mined ore or stockpiled ore. Vale’s announced timing of the Salobo III expansion is to have first production in January 2022. Assuming the Salobo III expansion achieves 12 Mtpa of additional processing capacity (bringing total processing capacity of the Salobo Mine to 36 Mtpa) and the current mine plan as shown in Section 16.2 (which achieves the higher plant throughput by processing ore that would have been stockpiled in the 24 Mtpa production plan), Wheaton would expect to pay an estimated Expansion Payment of approximately $570 million if completion is achieved in 2022 or $550 million if achieved in 2023. However, should Vale adopt a mine plan which maintains the current elevated copper cut-off grade by the time of completion of the Salobo III expansion, Wheaton’s estimated Expansion Payment would be approximately $670 million in 2022 or $650 million if achieved in 2023. The economics of the Salobo mine were evaluated using an after-tax discounted cash flow (DCF) model on 100% basis including the current processing lines plus the Salobo III expansion. The following sections detail the assumptions and results of the analysis.

22.1. Key Assumptions The following are the key assumptions used in the Salobo economic analysis:

• The financial calculations are based on an after-tax discount rate of 6.3%. All costs and prices are not escalated “real” dollars. Taxes are approximated with a 34% federal tax rate, being 25% income tax rate and 9% of social contribution.

December 2019 148

However, it is assumed SUDAM benefit, a legal Amazon area incentive, which represents a 75% reduction in income tax rate (from 25% to 6.25%) until 2023.

• All costs / investments are based on the 5-year plan budget version as of late October 2019, (assuming exchange rate of 3.80 BRL/US$). Post 2024, costs were estimated based on the 2020-2024 period.

• Closure costs were considered in the model totaling approximately $194 M.

• Revenue is calculated from the recoverable metals and the long-term forecast of metal prices and exchange rate.

• Debentures payment to the Brazilian Development Bank (BNDES) are included in the results.

• Reduction in Au revenue as a result of the Wheaton streaming deals is considered in the results as well as Wheaton’s Expansion Payment to be made in 2022.

Economic evaluations determine viability of the operation and include the following assumptions:

• The economic viability for both Mineral Reserves and Mineral Resources is demonstrated by the evaluation of the revenue generated from the long-term production plan against all applicable costs.

• All assumptions supporting the Mineral Reserve estimates (e.g. metal sale price assumptions, exchange rate assumptions, operating costs, capital costs, discount rates, etc.) are updated at least once a year.

• Site closure costs of approximately $194 M that include decommissioning of the open pit, waste dumps, tailings dam, industrial facilities and site infrastructure.

• A 2% CFEM – Federal Royalty – Mineral Exploitation Levy payable to the Brazilian Government.

• Mineral Resources are not used in support of the Mineral Reserves economic evaluation.

• For the mineralization that otherwise meets the definition of Mineral Reserves (and management intends to mine) economic viability is demonstrated when the projected discounted cash flows in a financial model return a positive Net Present Value (NPV).

Table 54 contains the metal price and exchange rates assumed annually until 2024 and then long term.

Table 54 - Metal Sale Price and Exchange Rate Assumptions

Item Units 2020 2021 2022 2023 2024 Long Term

Copper Price US$/tonne

Cu 6,125 6,200 6,250 6,600 7,100 7,000

Gold Price US$/oz. Au 1,325 1,350 1,350 1,370 1,360 1,290 Exchange

Rate BRL/US$ 3.8 3.8 3.8 3.8 3.8 3.8

December 2019 149

22.2. Cash Flow Analysis The financial model that supports the mineral reserve declaration is a standalone model that calculates annual cash flows based on scheduled ore production, assumed processing recoveries, metal sale prices and BRL/US$ exchange rate, projected operating and capital costs and estimated taxes. Table 55 summarizes the results of the cash flow analysis for the Salobo Operations 2019 MRMR Statement and Table 56 shows the annual free cash flows. This cash flow is based upon the mineral reserve estimate for Salobo Operations as reflected in this report, representing an approximate 33-year life. The Salobo Operations total cash flow is estimated to be US$9,859 M from 2020 to the end of the estimated mineral reserve life, including closure costs. The total mine production included in the financial model is 1,148 M tonnes grading 0.60% Cu and 0.32 g/t Au. The average annual copper production from 2020 to 2052 is 180,600 tonnes.

Table 55 - Cash Flow (Reserves only)

Cash Flow Item Total LoMP

(US$ M)

Net Revenue 45,734

Total Operating Costs -21,481

Corporate Overhead -322

Pre-operational Expenses -13

Other expenses/income -79

R&D -19

EBITDA 23,821

Capital Investment -969

Sustaining capital -2,193

Working capital changes 94

Net Cash Flow before Taxes 20,752

Taxes -5,864

Mine closure costs -194

Debentures to BNDES -200

Au streaming impact -4,635

Total Cash Flow 9,859

December 2019 150

Table 56 - Annual Free Cash Flow (Reserves Only)

Year Free Cash Flow (US$M) Year Free Cash Flow (US$M)

2020 -132.4 2037 270.6

2021 -146.5 2038 284.0

2022 820.4 2039 239.3

2023 564.5 2040 464.1

2024 426.4 2041 569.0

2025 370.2 2042 538.7

2026 345.7 2043 443.3

2027 315.0 2044 478.8

2028 317.1 2045 242.5

2029 346.8 2046 158.9

2030 354.0 2047 155.2

2031 361.9 2048 155.4

2032 297.7 2049 150.3

2033 404.6 2050 148.5

2034 297.6 2051 146.9

2035 300.2 2052 49.8

2036 273.4 2053+ -152.9

22.3. DCF Analysis Results and Sensitivity Analysis Table 57 shows the net present value (NPV) of the Salobo mine assuming a discount rate of 6.3%.

Table 57 - NPV – Mineral Reserves Only

Discount Rate

NPV (US$ M)

6.3% 4,413

The spider diagram in Figure 55 shows the sensitivity of the Salobo mine with changes in copper and gold price, OPEX, CAPEX and exchange rate ranging from negative 20% to positive 20%. The diagram shows that copper price is the most sensitive item to NPV, followed by OPEX and exchange rate. Due to the Wheaton streaming deals, gold price has very little effect on NPV.

December 2019 151

Figure 55 - NPV Sensitivity Analysis

22.4. Comments on Section 22 In the QPs opinion, Vale’s annual cash flow analysis in support of Mineral Reserve declaration is robust and shows the strong economics of the Salobo mine. The Mineral Resources contain additional value which was not considered in this analysis.

23. Adjacent Properties This section is not relevant to this Report.

24. Other Relevant Information There is no other relevant data and information to disclose.

25. Interpretation and Conclusions The Salobo Operations have been in operation since 2009 (pre-stripping) and has successfully ramped up to design capacity of 24 Mtpa and designed process recoveries. Vale operates the mine according to high standards with respect to safety, operating practices and the environment. Construction of the Salobo III expansion began in 2019 and will add another 12 Mtpa of process plant capacity. First production from Salobo III is expected in January 2022. Mineral Resources and Mineral Reserves have been prepared according to the 2014 CIM Definition Standards and this Report according to the requirements of National Instrument 43-101 Standards of Disclosure for Mineral Projects and Form 43–101F1.

1,000

2,000

3,000

4,000

5,000

6,000

7,000

-20% -10% 0% 10% 20%

NP

V @

6.3

%

Cu price

Au price

Opex

Capex

Exchange Rate

December 2019 152

The mine has a large Mineral Reserve base and strong economic margins which result in forecasted mining until the year 2044 and then the processing of stockpiled material until 2052. In order to maintain strong operating performance, the established production reconciliation control and operational mining dilution studies must be continued (selective mining unit and equipment operations) to ensure that best criteria are being applied to the Mineral Resources and Mineral Reserves definition. The areas that have the highest impact on production are mining dilution and recovery, as well as, the challenge to optimize mining operations. Risks include meeting the material movement targets and drilling/production equipment efficiency.

26. Recommendations At all mining operations there are areas to improve efficiencies and reduce costs. The following recommendations are presented as items that could have a positive impact to the mine economics:

1. Exploration drilling below Salobo final pit has shown that the deposit has continuity. More drilling is required to evaluate how extensive the deposit is below the pit.

2. The current Resource pit bottoms out on the bottom of the block model. Now that

the deep drilling has confirmed mineralization below that level the block model base should be lowered during the next update.

3. The Reserve pit has not been updated since 2016. A new Reserve pit optimization should be done in 2020 on the updated Resource model. The extended depth of the Resource model could result in an increased Reserve pit which could affect the cut-off strategy currently envisioned for the Salobo III expansion.

4. Review of the operational pit slope angles through geotechnical examination of the pit wall operation, design of pushbacks, and further geotechnical studies may provide support for steepening of some of the pit walls. The current geomechanical sectors are based on a limited amount of information and this should be further analyzed.

5. Salobo Operations should keep the established production reconciliation control and operational mining dilution studies current (selectivity mining unit and equipment operations) to ensure that best criteria are being applied to the Mineral Resources and Mineral Reserves definition. Continuous improvements in the ore dilution control and follow up process will result in improvement of the F2 reconciliation (mining recovery and dilution).

December 2019 153

6. Infill drilling program has shown some unexpected geological results and several zones with the low drilling density represent opportunities for further investigation.

7. The QPs agree with the ramp-up assumptions for the Salobo III line of the process plant and the use of dedicated metallurgical samplers in the reject and concentrate lines of the plant and the utilization of the primary crusher weighing scales to evaluate the dispatch system truck data measurement and calibration are of high importance. This should result in an improvement in the metallurgical balance and higher confidence in the reconciliation factors.

27. References AmecFW, 2016: Salobo Operations Pará State, Brazil NI 43-101 Report: technical report prepared by AmecFW for Wheaton, effective date 31 December 2015. Burns, N., Davis, C., Diedrich, C., Tagami, M., 2018: Salobo Copper-Gold Mine Carajas, Para State, Brazil Technical Report, effective date 31 December 2017. BVP Engenharia, 2007: Relatório de Escavabilidade Projeto Salobo: report prepared for Companhia Vale do Rio Doce. BVP Engenharia, 2007: Revisão dos taludes da cava: report prepared for Companhia Vale do Rio Doce. Canadian Institute of Mining, Metallurgy and Petroleum (CIM), 2019: Estimation of Mineral Resources and Mineral Reserves, Best Practice Guidelines: Canadian Institute of Mining, Metallurgy and Petroleum, November 29, 2019. Canadian Institute of Mining, Metallurgy and Petroleum (CIM), 2014: CIM Standards for Mineral Resources and Mineral Reserves, Definitions and Guidelines: Canadian Institute of Mining, Metallurgy and Petroleum, May, 2014. Canadian Securities Administrators (CSA), 2011: National Instrument 43-101, Standards of Disclosure for Mineral Projects, Canadian Securities Administrators. Companhia Vale do Rio Doce, 2003: Projeto Salobo. Validação de Resultados Analíticos de Projeto: internal report prepared by Companhia Vale do Rio Doce, September 2003. Companhia Vale do Rio Doce, 2006: Plano De Manejo Para Uso Múltiplo Da Floresta Nacional Do Tapirapé-Aquiri: report prepared by Companhia Vale do Rio Doce, September, 2006. Companhia Vale do Rio Doce and Salobo Metais S.A., 2003a: Salobo 3 Alfa Deposit Geological Model: internal PowerPoint presentation prepared by Companhia Vale do Rio Doce and Salobo Metais S.A.

December 2019 154

Companhia Vale do Rio Doce and Salobo Metais S.A., 2003b: Relatório Final do Depósito Salobo 3 Alfa: internal report prepared by Companhia Vale do Rio Doce and Salobo Metais S.A. De Souza, L. H. and Vieira, E. A., 1998: Jazida de Cobre Salobo/Ouro/Prata, Pará, Brasil, Relatório Final de Geologia, Fase Salobo Metais S.A. (1997–1998): internal report prepared by Salobo Metais S.A., Rio de Janeiro, April, 1998. Equipe de Geometalurgia, 2015: Projeto de Geometalurgia, Mina do Salobo: report prepared by Vale, May 2015. Golder Associates, 2010: Salobo Copper Gold External Resource and Reserves Audit 2011: report prepared for Vale. Matos da Costa, M.A., 2012: Proveniência Dos Arenitos da Formação Águas Claras, Província Mineral de Carajás, Se do Cráton Amazônico: M.Sc. Thesis, Instituto de Geociência, Universidade Federal do Pará. Micon International Limited, 2013: Technical Report on the Mineral Reserves and Mineral Resources of the Salobo Copper-Gold Mine Carajás, Pará State, Brazil: technical report prepared by Micon for Silver Wheaton Corp., effective date 31 December, 2012. Pincock, Allen and Holt, 2007: Salobo Copper Gold Project, Full Due Diligence Audit, Pará State, Brazil: report prepared for CVRD, 9 March, 2007. Reuwsaat, J.D.V., Ribeiro, C., and de Paula, W.F., 2014: Vale Base Metals, South Atlantic Salobo Operations, Carajás Mining District, Pará State, Brazil, Mineral Resource and Mineral Reserve Estimate, 2014 Technical Report: internal Vale report, 31 December, 2014. Snowden, 2009: Salobo Copper Gold External Resource Audit 2009: report prepared for CVRD. Stacey Mining Geotechnical Ltd., 2009: Mining Geotechnical Review, report prepared for Vale, document DIOC LR 01-09. U.I. Minerals, 2004: Consolidation of Data On Variability Rougher Campaign Locked-Cycle-Tests, Mini-Pilot-Plant, 1994 Pilot Plant Campaign, and Proposed Process Criteria For Salobo Flotation Plant: internal report prepared by U.I. Minerals, revision 2, February 2004. Vale, 2018: Vale Base Metals Guidelines and Standards for Mineral Resource Mineral Reserve Reporting: internal Vale document prepared by the Resource Management Group, revised 29 March, 2018.

December 2019 155

Vale, 2018. Periodic Report QA/QC – Chemical Data Drilling Campaign 2017/2018 Salobo Mine: internal Vale document prepared by the Exploration and Mineral Projects groups, August 2018.

Salobo Copper-Gold Mine Salobo III Expansion · assessment, pre-feasibility, and feasibility studies, and due diligence studies in Canada, Brazil, Tanzania, Ghana and Portugal. As

Documents